38

PHARMACOKINETICS,
PHARMACODYNAMICS, AND DRUG
DISPOSITION
DAVID J. GREENBLATT
LISA L. VON MOLTKE
JEROLD S. HARMATZ
RICHARD I. SHADER

During the last decade, the application of pharmacokinetic cokinetic models are applied to determine parameters such
and pharmacodynamic modeling techniques has become an as elimination half-life, volume of distribution, and clear-
increasingly important aspect of contemporary clinical psy- ance. During the new drug development process, a series
chopharmacology (15). These techniques have been ap- of pharmacokinetic studies are conducted to determine the
plied during the process of development of new drug entities influence of major disease states or experimental conditions
as well as for the improved understanding of the clinical hypothesized to affect drug disposition. Such factors might
actions of drugs that are already marketed. Techniques for include age, gender, body weight, ethnicity, hepatic and
the study of drug metabolism in vitro have advanced sub- renal disease, coadministration of food, and various drug
stantially during the last decade, and now are an integral interactions. Classical pharmacokinetic studies can quanti-
component of preclinical drug development and the link tate the effects of anticipated influences on drug disposition
to subsequent clinical studies of drug metabolism and dispo- under controlled circumstances, but cannot identify the un-
sition. Kinetic-dynamic modeling techniques have been expected factors affecting pharmacokinetics. A number of
combined with in vitro metabolism procedures and in vi- examples of altered drug pharmacokinetics became apparent
troin vivo mathematical scaling models to provide insight in the patient care setting only in the postmarketing phase
into the general problem of pharmacokinetic drug interac- of extensive clinical use. Examples include the digoxin-quin-
tions in clinical psychopharmacology (69). idine interaction, altered drug metabolism due to cimeti-
This chapter reviews some advances in pharmacokinetics, dine, and the ketoconazole-terfenadine interaction.
pharmacodynamics, and drug metabolism, along with Population pharmacokinetic methodology has developed
methodologic applications to selected problems in clinical
as an approach to detect and quantify unexpected influences
psychopharmacology.
on drug pharmacokinetics (1018). Population pharmaco-
kinetic studies, in contrast to classical or traditional pharma-
cokinetic studies, focus on the central tendency of a phar-
POPULATION PHARMACOKINETICS macokinetic parameter across an entire population, and
Principles identify deviations from that central tendency in a subgroup
of individual patients. One software program widely applied
Pharmacokinetic studies based on a traditional intensive- to population pharmacokinetic problems is the nonlinear
design model are usually conducted using carefully selected mixed-effects model (NONMEM). Analysis of clinical data
volunteer subjects, a controlled experimental design, and using a population approach allows pharmacokinetic pa-
collection of multiple blood samples. After measurement of rameters to be determined directly in patient populations
drug and metabolite concentrations in all samples, pharma- of interest and allows evaluation of the influence of various
patient characteristics on pharmacokinetics. Because the
number of blood samples that need to be collected per sub-
D. J. Greenblatt, L. L. von Moltke, J. S. Harmatz, and R. I. Shader: ject is small, this approach is often suitable for patient
Department of Pharmacology and Experimental Therapeutics, Tufts Univer-
sity School of Medicine, and Division of Clinical Pharmacology, New England groups unable to participate in traditional pharmacokinetic
Medical Center, Boston, Massachusetts. studies requiring multiple blood samples (e.g., neonates,
508 Neuropsychopharmacology: The Fifth Generation of Progress

children, critically ill patients, or individuals who are not

able to provide informed consent) (19). In many cases the
population approach has yielded pharmacokinetic parame-
ter estimates similar to those delineated in classical pharma-
cokinetic studies of the same drug.

Application: Methylphenidate
Pharmacokinetics
The population approach is illustrated in a study of methyl-
phenidate (MP) pharmacokinetics in children (20). This is
a patient group for whom the multiple-sample pharmacoki-
netic study design may not be appropriate for ethical and
practical reasons. Participating subjects were 273 children
aged 5 to 18 years having a primary diagnosis of attention- FIGURE 38.1. Population pharmacokinetic model for methyl-
deficit/hyperactivity disorder (ADHD). They had been re- phenidate (MP). A series of data points, each consisting of the
ceiving MP at a fixed dosage level for at least 4 weeks, and time (t) after the first dose of the day and the plasma MP concen-
tration (C) at that time, was available from 273 subjects (one data
were under treatment for at least 3 months. The treating point per subject). Each of these was linked to that subjects indi-
physician for each patient judged MP to be clinically effec- vidual dose schedule, size of each dose, interval between doses,
tive. and body weight. These variables were entered into a one-com-
partment pharmacokinetic model with first-order absorption and
Children meeting the eligibility criteria had an initial first-order elimination, as shown. Using nonlinear regression, the
screening visit, at which one parent or a legal guardian pro- process yielded typical population values of clearance per kilo-
vided written informed consent, and the child provided as- gram body weight, the elimination rate constant (Ke), and the
absorption rate constant (Ka).
sent. Demographic characteristics were recorded, including
the dosage of MP, the usual times for individual doses, and
the duration of treatment.
The second visit, which followed shortly, was a blood- variables using unweighted nonlinear regression (Fig. 38.1).
sampling day. Each child, accompanied by parent or guard- When the time between first and second doses, or between
ian, arrived at the investigators office 30 to 60 minutes second and third doses, was not available, the mean value
prior to blood sampling. The time and size of the last MP was assigned based on cases in which the data were available.
dose, and of any other medication received that day or dur- For the b.i.d. dosage, the mean interval was 4.3 hours. For
ing the prior 2 weeks, were recorded. A 5-mL whole blood the t.i.d. dosage, the mean intervals were 4.1 and 3.7 hours,
sample was obtained by venipuncture. This sample was im- respectively. As is customary, clearance was assumed to be
mediately centrifuged, and a 2-mL aliquot of plasma was proportional to body weight.
removed for subsequent determination of MP concentra-
tions by a liquid chromatography/mass spectroscopy/mass
Results
spectroscopy (LC/MS/MS) assay.
The total daily dose of MP was significantly lower in sub-
jects receiving MP b.i.d. (n 109) compared to subjects
Analysis of Data
on a t.i.d. schedule (n 164); the mean total daily dosages
The identified independent variables were age, sex, body in the two groups were 25 and 39.3 mg, respectively
weight, size of each dose, and time of sample relative to the (p .001). Within each group, clinicians choices of total
most recent dose. Since only single samples were available daily dosages were influenced by body weight, as mean total
for all but 16 of these children, the contribution of within- daily dose increased significantly with higher body weights.
subject variance to overall variability in outcome could not However, the association of body weight with mean plasma
be assessed. The pharmacokinetic model was a one-com- concentration was not significant for the b.i.d. dosage
partment model with first-order absorption and first-order group, and of only borderline significance (.05 p .1)
elimination, under the assumption that all subjects were at for the t.i.d group. This finding is consistent with the under-
steady state (Fig. 38.1). lying assumption that clearance is proportional to body
The overall model was specifically modified for each of weight.
the 273 subjects to incorporate the individually applicable Age was significantly correlated with body weight (r2
independent variables, as well as the dosage schedule (b.i.d. 0.54, p .001) and with height (r2 0.77). Height and
or t.i.d.). Individual values of continuous variables (t body weight also were significantly correlated (r2 0.77).
time sample taken relative to the first dose; C plasma An acceptable estimate of absorption rate constant could
MP concentration) were fitted to a single set of iterated be derived only for the b.i.d. dosing data. The iterated pa-
 38: Pharmacokinetics, Pharmacodynamics, and Drug Disposition 509

FIGURE 38.2. Overall relation of observed and predicted plasma FIGURE 38.3. Predicted plasma methylphenidate concentration
methylphenidate concentrations (ng/ml). The r-square value of curves for b.i.d. and t.i.d. dosage schedules, based on parameter
0.43 indicates that the model accounts for 43% of the overall estimates from the population analysis, together with mean val-
variance in plasma concentrations. (From Shader RI, Harmatz JS, ues of input variables (body weight, size of doses, intervals be-
Oesterheld JR, et al. Population pharmacokinetics of methylphen- tween doses). (From Shader RI, Harmatz JS, Oesterheld JR, et al.
idate in children with attention-deficit hyperactivity disorder. J Population pharmacokinetics of methylphenidate in children
Clin Pharmacol 1999;39:775785, with permission.) with attention-deficit hyperactivity disorder. J Clin Pharmacol
1999;39:775785, with permission).

rameter estimate was 1.192/h, corresponding to an absorp-

This may reflect the clinicians considering body weight in
tion half-life of 34.9 minutes. This estimate was then fixed,
their choice of total daily dosage, or it may be that the
and the entire data set analyzed to determine clearance per
dose was titrated according to response, which in turn was
kilogram of body weight, and the first-order elimination
influenced by associations among concentration, clearance,
rate constant. The iterated estimates were 0.154/h for elimi-
and weight.
nation rate constant, corresponding to an elimination half-
The pharmacokinetic model explained 43% of the vari-
life of 4.5 hours (relative standard error: 23%). For clear-
ability in plasma MP concentrations during typical natural-
ance, the estimate was 90.7 mL/min/kg (relative standard
istic therapy. The model fit equally well for both genders.
error: 9%). The overall r-square was 0.43 (Fig. 38.2). There
Assuming that clearance is proportional to body weight in
were no evident differences in pharmacokinetics attributable
the context of intercorrelated age and weight allows age,
to gender. Figure 38.3 shows predicted plasma MP concen-
weight, and daily dosage to be used to predict plasma con-
tration curves for b.i.d. and t.i.d. dosage schedules, based
centrations of MP during clinical use in children. These
on the population estimates.
findings support the value of prescribing MP on a weight-
adjusted basis.
Our typical population value of elimination half-life was
Implications
4.5 hours, with a confidence interval of 3.1 to 8.1 hours.
Pharmacokinetically based approaches to the treatment of This estimate somewhat exceeds the usual range of half-life
ADHD with MP are not clearly established (2125). In the values reported in single-dose kinetic studies of MP (25,
present study of prescribing patterns in particular clinical 26). This could reflect the relatively small number of plasma
practices, the mean prescribed per dose amount for the samples from the terminal phase of the plasma concentra-
whole study population was 0.335 mg/kg per dose (range tion curve, upon which reliable estimates of beta are depen-
0.0440.568), and 36% of the children received be- dent. MP kinetics may also have a previously unrecognized
tween 0.25 and 0.35 mg/kg per dose. The mean total daily dose-dependent component, in which estimated values of
dose was 0.98 mg/kg/day for the entire sample, and in- half-life are larger at steady state than following a single
creased significantly in association with larger body weight. dose.
510 Neuropsychopharmacology: The Fifth Generation of Progress

The single-sample approach described in this study al- necessarily depend on the type of drug under study. For
lows relatively noninvasive assessment of pharmacokinetic sedative-anxiolytic drugs such as benzodiazepines, effects of
parameters in a group of children and adolescents under interest may include subjective or observer ratings of seda-
naturalistic circumstances of usual clinical use, when blood tion and mood; semiobjective measures of psychomotor
sampling is not otherwise clinically indicated. This ap- performance, reaction time, or memory; or objective effect
proach in general can be applied to other special populations measures such as the EEG or saccadic eye movement veloc-
such as neonates, the elderly, or individuals with serious ity. The various measures differ substantially in their rele-
medical disease. vance to the principal therapeutic actions of the drug, the
stability of the measure in terms of response to placebo or
changes caused by practice or adaptation, the objective or
KINETIC-DYNAMIC MODELING subjective nature of the quantitative assessment, and the
comparability of results across different investigators and
Principles different laboratories (Table 38.1). The extent to which the
Pharmacokinetics is the discipline that applies mathematical various pharmacodynamic measures provide unique infor-
models to describe and predict the time course of drug con- mation, as opposed to being overlapping or redundant, is
centrations in body fluids, whereas pharmacodynamics re- not clearly established.
fers to the time course and intensity of drug effects on the Pharmacokinetic and pharmacodynamic relationships
organism, whether human or experimental animal (Fig. initially are evaluated separately, and the relationship of ef-
38.4). Both have evolved as the techniques for measuring fect versus concentration at corresponding times is exam-
drug concentrations, and drug effects have become more ined graphically and mathematically. Effect measures are
accurate and sensitive. Evolving in parallel is kinetic-dy- usually expressed as change scores: the net effect (E) at post-
namic modeling, in which the variable of time is incorpo- dosage time t is calculated as the absolute effect at this time
rated into the relationship of effect to concentration (Fig. (Et) minus the predose baseline value (Eo), that is, E
38.4) (2732). A concentration-effect relationship is, in Et Eo. Several mathematical relationships between effect
principle, the most clinically relevant, because it potentially and concentration (E versus C), often termed link models,
validates the clinical rationale for measuring drug concentra- are of theoretical and practical importance (5,32). The sig-
tions in serum or plasma. moid Emax model, incorporates a value of Emax, the maxi-
A kinetic-dynamic study in clinical psychopharmacology mum pharmacodynamic effect, and EC50 is the 50% effec-
typically involves medication administration (usually under tive concentration, the concentration that is associated
placebo-controlled, double-blind laboratory conditions) fol- with half of the maximum effect (Fig. 38.5). The exponent
lowed by quantitation of both drug concentration and clini- A reflects the steepness of the concentration-response re-
cal effect at multiple times after dosing. Measures of effect lationship in its ascending portion. The biological impor-
tance of A is not established.
A concentration-effect relationship that is consistent with
the sigmoid Emax model may be of mechanistic importance,
because drug-receptor interactions often fit the same model.
The Emax and EC50 values allow inferences about questions
such as the relative potency or efficacy of drugs producing
the same clinical effect, individual differences in drug sensi-
tivity, the mechanism of action of pharmacologic potentia-
tors or antagonists, and the possible clinical role of new
medications.
The sigmoid Emax model does not necessarily apply to
all concentration-effect data (32). When experimental data
are not consistent with the model, the corresponding misap-
plication of the sigmoid Emax relationship can lead to mis-
leading conclusions about Emax and EC50. Some data sets
are consistent with less complex models, such as exponential
or linear equations (Fig. 38.5); in these cases, the concepts
FIGURE 38.4. Schematic relation between pharmacokinetics,
of Emax and EC50 are not applicable. Kinetic-dynamic mod-
pharmacodynamics, and kinetic-dynamic modeling, based on the eling is further complicated when drug concentrations mea-
status of the variables of time (t), concentration (C), and effect sured in serum or plasma do not reflect the concentration
(E). Note that kinetic-dynamic modeling incorporates both phar-
macokinetics and pharmacodynamics, with time subsumed into at the site of action, which is sometimes termed the effect
the relation of concentration and effect. site. This is illustrated by the data described below.
 38: Pharmacokinetics, Pharmacodynamics, and Drug Disposition 511

TABLE 38.1. PHARMACODYNAMIC ENDPOINTS APPLICABLE TO STUDIES OF GABA-BENZODIAZEPINE

AGONISTS

Classification Relation to Primary Effect of Effect of Need for "Blind" Approach to

(with Examples) Therapeutic Action Placebo Adaptation/Practice Conditions Quantitation

Subjective
Global clinical ratings; Close Yes Yes Yes Transformation of
targeted rating scales ratings into numbers
Semi-objective
Psychomotor function May be linked to Yes Yes Yes Test outcomes are
tests; memory tests adverse effect quantitative
profile
Objective
Electroencephalography Not established No No No Fully objective
computer-determined
quantitation

GABA, -aminobutyric acid.

Application: Kinetics And Dynamics Of EEG was used as the principal pharmacodynamic outcome
Intravenous Lorazepam measure (Table 38.1). The EEG was recorded prior to lora-
zepam administration, and at times corresponding to blood
In this study the benzodiazepine derivative lorazepam was samples. EEG data were digitized over the power spectrum
administered intravenously according to a complex bolus- from 4 to 30 cycles per second (Hz), and analyzed by fast
infusion scheme (33). On the morning of the study day, a Fourier transform to determine amplitude in the total spec-
rapid intravenous dose of lorazepam, 2 mg, was adminis- trum (4 to 30 Hz) and in the beta (12 to 30 Hz) frequency
tered into an antecubital vein, coincident with the start of range (3335). Concentrations of lorazepam in plasma sam-
a zero-order infusion at a rate of 2 g/kg/h. The infusion ples were determined by gas-chromatography with electron-
continued for 4 hours and then was terminated. Venous capture detection.
blood samples were drawn from the arm contralateral to
the site of the infusion prior to drug administration and
at multiple time points during 24 hours after the start of Analysis of Data
lorazepam infusion. Samples were centrifuged, and the The relative EEG beta amplitudes (beta divided by total,
plasma separated and frozen until the time of assay. The expressed as percent) in the predose recordings were used
as the baseline. All values after lorazepam administration
were expressed as the increment or decrement over the mean
predose baseline value, with values averaged across eight
recording sites. The EEG change values were subsequently
used as pharmacodynamic effect (E) measures in kinetic-
dynamic modeling procedures described below. For phar-
macokinetic modeling, the relation of plasma lorazepam
concentration (C) to time (t) was assumed to be consistent
with a two-compartment model (Figs. 38.6 and 38.7).
Examination of plots of pharmacodynamic EEG effect
versus plasma lorazepam concentration (E vs. C) indicated
counterclockwise hysteresis (see below), suggesting a delay
in equilibration of lorazepam between plasma and the site
of pharmacodynamic action in brain. This equilibration ef-
fect has been described in previous clinical and experimental
studies of lorazepam (34,3639). Accordingly the relation-
FIGURE 38.5. Three mathematical relationships between con-
centration (C) and change in pharmacodynamic effect (E) that are ship was modified to incorporate a distinct effect site, at
commonly applied in kinetic-dynamic modeling procedures. For which the hypothetical lorazepam concentration is CE (Fig.
the sigmoid Emax model, Emax is maximum pharmacodynamic ef- 38.6). The apparent rate constant for drug disappearance
fect, EC50 is the concentration producing a value of E equal to
50% of Emax, and A is an exponent. For the exponential and linear from the effect compartment is kEO; this rate constant deter-
models, m is a slope factor. mines the apparent half-life of drug equilibration between
512 Neuropsychopharmacology: The Fifth Generation of Progress

plasma and effect site. Under these assumptions, the relation

of E to CE was postulated to be consistent with a sigmoid
Emax model (Fig. 38.5).

Results
Kinetic variables for lorazepam were similar to those re-
ported in previous single-dose studies of lorazepam pharma-
cokinetics (34,4045). Overall mean values were volume
of distribution, 1.7 L/kg; elimination half-life, 14 hours;
clearance, 1.44 mL/min/kg. The bolus-infusion scheme rap-
idly produced mean plasma lorazepam concentrations in
the range of 18 to 19 ng/mL, values close to the mean
predicted value of 24 ng/mL.
FIGURE 38.6. Schematic representation of the kinetic-dynamic Lorazepam infusion produced significant increases in
model for the lorazepam study. Intravenous lorazepam was as-
sumed to have kinetic behavior consistent with a two-compart-
EEG beta amplitude throughout the 24-hour duration of
ment model: reversible distribution to a peripheral compartment, the study. The maximum change over baseline was mea-
and first-order elimination (clearance) from the central compart- sured at 0.25 to 0.75 hours after the initiation of lorazepam
ment (rate constant: Ke). Lorazepam in plasma was postulated to
equilibrate with a hypothetical effect site, from which the exit
dosage, whereas the maximum plasma concentration was
rate constant is KEO. Finally, effect-site concentrations were pre- measured immediately after the loading dose (Fig. 38.8).
sumed to be the principal determinant of pharmacodynamic ef- The effect-site model eliminated the hysteresis, with a mean
fect, via a kinetic-dynamic link model as shown in Fig. 38.5.
equilibration half-life of 8.8 minutes (Fig. 38.9).

Implications
Maximum EEG effects of lorazepam were significantly de-
layed following the initial intravenous bolus dose. Previous
single-dose pharmacodynamic studies of lorazepam, using
the EEG or other methods for quantitation of benzodiaze-

FIGURE 38.7. Plasma lorazepam concentrations (solid circles) to-

gether with the pharmacokinetic function determined by nonlin- FIGURE 38.8. Mean values of plasma lorazepam concentration,
ear regression (solid line), in a representative volunteer subject. and of EEG beta amplitude, during the first hour of the study.
Shown are the derived pharmacokinetic variables of elimination Note that pharmacodynamic EEG effects are delayed following
half-life (t1/2), volume of distribution (Vd), and clearance (CL). the peak value of lorazepam in plasma.
 38: Pharmacokinetics, Pharmacodynamics, and Drug Disposition 513

FIGURE 38.9. Left: Mean values of plasma lorazepam concentration versus pharmacodynamic
EEG effect at corresponding times, with arrows indicating the direction of increasing time. As
indicated in Fig. 38.8, the maximum EEG effect is delayed, and does not correspond in time to
the maximum plasma concentration. Right: The scheme shown in Fig. 38.6 was applied to the
data points, with the link model being the sigmoid Emax relationship shown in Fig. 38.5. The data
points (closed triangles) are the hypothetical effect site concentrations and pharmacodynamic
effect values at corresponding times. The solid line is the link model function determined by
nonlinear regression, yielding the indicated values of Emax and EC50. The overall mean equilibration
half-life was 8.8 minutes.

pine effect, consistently demonstrate a delay in attainment tance for the metabolism and clearance of most drugs used
of maximum drug effect compared to attainment of peak in psychopharmacology and in other areas of clinical thera-
concentrations in plasma (34,36,37). After rapid intrave- peutics (69,4855) (Fig. 38.10). For the CYP isoforms
nous dosage, for example, maximum effects may be delayed most relevant to human drug metabolism, each has its own
for an average of 30 minutes after dosage. Experimental distinct pattern of relative abundance, anatomic location,
studies of the time-course of whole-brain concentrations mechanism of regulation, substrate specificity, and suscepti-
of lorazepam, or of the degree of benzodiazepine receptor bility to inhibition and induction by other drugs or foreign
occupancy, indicate that the delay is attributable to the slow chemicals (Table 38.2). The expression and in vivo function
physical entry of lorazepam into brain tissue, probably be- of at least two CYP isoforms (CYP2D6 and CYP2C19)
cause of the relatively low lipid solubility of lorazepam (34, are regulated by a genetic polymorphism, such that some
38,39). The delay was mathematically consistent with a ki- members of a population fail to express normal levels
netic-dynamic model incorporating a hypothetical effect of enzyme or expresses poorly functional protein (5662).
site distinct from the central compartment. The half-life Individuals identified as CYP2D6 poor metabolizers, as
for equilibration between plasma and the effect compart- an example, have very low clearance of drugs that are major
ment was approximately 9 minutes. This matches clinical substrates for biotransformation by CYP2D6 (such as desi-
experience indicating that intravenous lorazepam cannot pramine, nortriptyline, venlafaxine, tramadol, and dextro-
easily be used in situations requiring minute-to-minute ti- methorphan). Such individuals are at risk for developing
tration of sedative or amnestic effects (40). Nonetheless, high and potentially toxic plasma concentrations of these
intravenous lorazepam can be used for the treatment of sta-
tus epilepticus, although its onset of action may be some-
what slower than that of intravenous diazepam (46,47).

CYTOCHROMES P-450 IN
PSYCHOPHARMACOLOGY: THE
IMPORTANCE OF P-450-3A ISOFORMS
FIGURE 38.10. Nomenclature system for the cytochrome P-450
(CYP) superfamily of enzymes. Following the CYP designation, the
The cytochrome P-450 (CYP) superfamily of drug metabo- number-letter-number sequence indicates the family, subfamily,
lizing enzymes is now established as being of primary impor- and specific isoform.
514 Neuropsychopharmacology: The Fifth Generation of Progress

TABLE 38.2. OVERVIEW OF HUMAN CYTOCHROMES P-450

Relative Hepatic Genetic

CYP Isoform Abundance Polymorphism Representative Substrates

1A2 13% ? Caffeine, theophylline, tacrine

2B6 <1% Bupropion, propofol
2C9 15% + S-warfarin, phenytoin,
tolbutamide, NSAIDs
2C19 4% + S-mephenytoin, omeprazole
(partial contributor to many
others)
2D6 2% + Some psychotropic and cardiac
drugs
2E1 7% Chlorzoxazone, some inhaled
anesthetics
3A4/5 29%a Many (see also Table 38.3)

CYP, cytochrome P-450; NSAID, nonsteroidal antiinflammatory drug.

a
Also present in gastrointestinal tract mucosa.

substrate drugs despite dosages in the usual therapeutic in varying amounts in some human livers and in esophagus,
range. but quantities of CYP3A5 are less than quantities of
CYP3A4. It is not established to what extent hepatic
CYP3A5 is of clinical significance for drug-metabolizing
The CYP3A Isoforms activity. CYP3A7 is principally a fetal-specific isoform. The
The overall importance of the CYP3A subfamily of drug- location and sequence of the genetic element responsible
metabolizing enzymes, particularly in the field of psycho- for CYP3A4 expression have been identified, as well as a
pharmacology, has become increasingly evident over the last regulatory segment located on the 5 flanking region corre-
decade (69,6369) (Table 38.3). The CYP3A isoforms sponding to the CYP3A gene.
are the most abundant of the CYPs, accounting on average CYP3A4 typically functions as a high-capacity, low-af-
for approximately 29% of identified cytochrome P-450 in finity enzyme. Its high substrate capacity is a consequence of
liver (70) (Table 38.2). Within the CYP3A subfamily, both the relatively high value of maximum reaction velocity
CYP3A4 is the most important in the adult human, in terms (Vmax, expressed in nanomoles of product produced per unit
of drug-metabolizing activity as well as quantitative domi- time per milligram of protein) in a Michaelis-Menten rela-
nance. CYP3A5, another CYP3A isoform, is also detected tionship, as well as the high quantitative abundance of the
protein in hepatic tissue. The low-affinity characteristic is
reflected in the high Km value (substrate concentration cor-
responding to 50% of Vmax) in a Michaelis-Menten relation-
TABLE 38.3. PSYCHOTROPIC DRUG SUBSTRATES ship. One consequence is that CYP3A-mediated metabo-
FOR HUMAN CYP3A lism usually is not saturable at substrate concentrations
within the therapeutic range, because this range is likely to
Contribution of CYP3A to Net Clearance
be far below the reaction Km. Furthermore, in situations in
Complete or which CYP3A is one of several cytochromes contributing
Nearly Complete Partial Small to metabolism [e.g., amitriptyline N-demethylation (71),
Midazolam Diazepam Fluoxetine
citalopram N-demethylation (72), or zolpidem hydroxyla-
Triazolam Desmethyldiazepam Sertraline tion (73)], the relative importance of CYP3A will increase
Alprazolam Flunitrazepam Nortriptyline at higher substrate concentrations. However, this is not in-
Bromazepam Clonazepam variably true. Nefazodone is a CYP3A substrate, but Km
Nefazodone Zolpidem
values for production of the various metabolites are rela-
Trazodone Citalopram
Reboxetine Haloperidol tively low (74), and kinetics are nonlinear (75). Midazolam
Buspirone Clozapine has a low Km for the principal pathway (76,77), and there
Gepirone Olanzapine is evidence of nonlinear kinetics at higher concentrations
Adinazolam Mirtazapine in humans (78).
Quetiapine Amitriptyline
Sildenafil Imipramine
Significant quantities of CYP3A exist in gastrointestinal
(GI) tract mucosa (65,69,79). The quantitative expression/
 38: Pharmacokinetics, Pharmacodynamics, and Drug Disposition 515

TABLE 38.4. REPRESENTATIVE DRUGS HAVING

CLINICALLY IMPORTANT EFFECTS ON THE
HUMAN CYP ENZYMES

Drug Inhibition of: Induction of:

Azole antifungals
Ketoconazole CYP3A
Itraconazole CYP3A
Fluconazole CYP3A, 2C9
Terbinafine CYP2D6
Antidepressants
Fluoxetine CYP2D6
Paroxetine CYP2D6
Fluvoxamine CYP1A2, 2C19, 3A
Nefazodone CYP3A
St. John's wort CYP3A
FIGURE 38.11. Relative contributions of CYP3A enzymes present Antipsychotics
in gastrointestinal (GI) tract mucosa, and in the liver, to net bio- Perphenazine CYP2D6
availability (F) of orally administered midazolam and triazolam. Anticonvulsants
Both of these compounds have net F values of less than 50% (29% Carbamazepine CYP3A
for midazolam, 44% for triazolam). Both compounds undergo Antithrombotics
approximately 50% extraction during passage through the G.I.
Ticlopidine CYP2D6, 2C19
tract mucosa. However midazolam undergoes another 38% ex-
traction across the liver, compared to only 12% for triazolam. Antiinfectives
Erythromycin CYP3A
Clarithromycin CYP3A
Ciprofloxacin CYP1A2
Rifampin CYP3A
activity of GI tract CYP3A is not correlated with its expres- Viral protease inhibitors
sion/activity in liver, even though the expressed protein is Ritonavir CYP3A CYP3A
identical at the two sites. For a number of moderate or high- Nonnucleoside reverse transcriptase inhibitors
Delavirdine CYP3A
clearance CYP3A substrates (e.g., midazolam and triazo-
Nevirapine CYP3A
lam), GI tract metabolism contributes importantly to pre- Cardiovascular agents
systemic extraction (first-pass metabolism) after oral dosage Quinidine CYP2D6
(7981); incomplete oral bioavailability therefore results Diltiazem CYP3A
from a combination of GI tract and hepatic presystemic Verapamil CYP3A
Antiulcer agents
extraction (Fig. 38.11). For low-clearance CYP3A substrates
Cimetidine CYP3A
having oral bioavailability in the range of 80% to 90% or Omeprazole CYP2C19
greater (e.g., alprazolam), the contribution of the GI tract
is apparently small.
Inhibition and induction by other drugs or chemicals
may modify CYP3A activity both in vitro and in vivo (Table
38.4). Identification of these compounds is of clear clinical cleoside reverse transcriptase inhibitor delavirdine also are
importance, because it may allow anticipating of drug inter- potent CYP3A inhibitors (8891). A component of grape-
actions that may be either potentially hazardous or of thera- fruit juice inhibits CYP3A in the GI tract (92). Inducers
peutic benefit (69,65,69,7985). Inhibiting drugs may of CYP3A include carbamazepine, rifampin, phenobarbital,
also be used for investigating the relative contribution of nevirapine, dexamethasone, St. Johns wort, and possibly
CYP3A to net clearance, or for distinguishing the contribu- venlafaxine. Ritonavir is an inducer as well as an inhibitor,
tion of hepatic and GI tract CYP3A to overall presystemic yielding a net effect on CYP3A metabolism that is difficult
extraction (81). Among the most potent CYP3A inhibitors to predict (8891,9395).
are the azole antifungal agents (ketoconazole, itraconazole, Variability among individuals in CYP3A activity is sub-
fluconazole), the antidepressants nefazodone and fluvoxa- stantial, even when relatively homogeneous groups of
mine, and the calcium channel antagonists verapamil and healthy subjects are studied. A consistent finding in popula-
diltiazem. These compounds produce reversible inhibi- tion studies of CYP3A activity is that distributions are uni-
tion, by a competitive, noncompetitive, or mixed mecha- modal, without evidence of genetic polymorphic regulation
nism. Other potent inhibitors, such as the macrolide anti- (96,97). However, several studies of CYP3A substrates have
microbials erythromycin and clarithromycin produce demonstrated a small number of individuals with unusually
mechanism-based inhibition via a metabolic intermediate low clearance (96,98). The explanation for these observa-
that complexes with and inactivates the CYP3A enzyme (86, tions is unclear, but the genomic determinants of such indi-
87). The HIV protease inhibitor ritonavir and the nonnu- vidual variations in clinical CYP3A activity have become
516 Neuropsychopharmacology: The Fifth Generation of Progress

FIGURE 38.12. Example of an in vitro metabolism study using human liver microsomes (73). The
substrate in this study was zolpidem, of which varying concentrations were incubated with liver
microsomes and appropriate reaction cofactors. At each concentration of zolpidem, the rate of
formation (V) of the principal metabolite of zolpidem (termed the M-3 metabolite) was deter-
mined. The relation between substrate concentration (S) and reaction velocity (V) was analyzed
by nonlinear regression to determine the maximum reaction velocity (Vmax) and the substrate
concentration (Km) producing a reaction velocity of 50% of Vmax. Left: A substrate concentration
range up to 2,000 M. Right: The lower range of concentrations shown on an expanded scale.

an active research topic. A number of genetic variants or inhibitors can also be used in clinical studies, but the in
single nucleotide polymorphisms in either the promoter or vitro model has the advantages of lower cost, more rapid
coding regions of the human CYP3A4 gene have recently implementation, no risk of human drug exposure, the avail-
been described (99105). One of the promoter region poly- ability of a greater number of potential chemical inhibitors
morphisms, designated as CYP3A4*1B, is more prevalent for research purposes, and the possibility of determining
in the African-American as opposed to the Caucasian popu- both the quantitative and qualitative contributions of spe-
lations. However, there is no evidence to indicate that any cific cytochromes. Antibodies with relatively specific inhibi-
of the identified CYP3A4 variants accounts for individual tory activity against individual human cytochromes can also
variation in clearance of CYP3A substrates. be used to support or confirm data from in vitro chemical
inhibition studies (106,107). In vitro approaches have been
strengthened with the availability of microsomes containing
In Vitro Models of Drug Metabolism pure human cytochromes as expressed by cDNA-transfected
In vitro systems now are extensively utilized to provide pre- human lymphoblastoid cells (108110). These heterolo-
sumptive answers to fundamental clinical questions regard- gously expressed pure cytochromes further support defini-
ing drug metabolism and drug interactions, and to guide the tive identification of cytochromes mediating a specific reac-
planning of clinical pharmacokinetic studies (69,4855, tion in vivo.
7174,8284). If drug X is biotransformed in humans to The quantitative inhibitory potency of a series of drugs
metabolite Y, two core questions occur: (a) What CYP iso- and their metabolites against specific index reactions can
form or isoforms mediate the biotransformation of X to Y? also be determined using human liver microsomes in vitro
(b) What CYP isoforms do X or Y themselves either induce (6,7). The first of two general approaches uses fixed concen-
or inhibit? trations of the index substrate co-incubated with varying
Human liver microsomes generally are an important concentrations of the inhibitor in question. The relation of
component of currently utilized in vitro systems. These decrement in metabolite formation rate to inhibitor concen-
preparations contain the various human CYPs in proportion tration is used to estimate a 50% inhibitory concentration
to their abundance in human liver in vivo. If biotransforma- (IC50). This procedure is expeditious and relatively inexpen-
tion of a specific substrate to its initial metabolite or metabo- sive, and the numbers can be used to compare the potency
lites can be replicated in microsomal preparations (Fig. of a series of inhibitors (Fig. 38.13). IC50 values themselves
38.12), inhibition of that reaction by a relatively specific are not dependent on knowledge of the specific biochemical
chemical inhibitor can be used as evidence supporting the mechanism of inhibition. However, IC50 values depend on
contribution of the corresponding cytochrome. Chemical substrate concentration when inhibition is competitive, and
 38: Pharmacokinetics, Pharmacodynamics, and Drug Disposition 517

though the relative rank order of inhibitory potency should

be maintained.
Limitations and drawbacks of in vitro systems should be
recognized. In vitro studies generally utilize substrate con-
centrations that are one or more orders of magnitude higher
than those encountered clinically, even if extensive parti-
tioning of lipophilic drugs from plasma into liver tissue
is accounted for. If mathematical models and parameter
estimates are valid, the outcome of studies of higher concen-
trations can be extrapolated down to clinically relevant sub-
strate concentrations. However, a clinically important
high-affinity metabolic reaction (i.e., one with a low Km
value) could be missed if low substrate concentrations can-
not be accurately measured due to limitations of assay sensi-
tivity. The specificity of chemical inhibitor probes is of con-
FIGURE 38.13. Example of an in vitro study of inhibition of cern for in vitro and in vivo studies, because all inhibitors
CYP2D6 by two antipsychotic agents, perphenazine and clozap- ultimately become nonspecific at higher concentrations. Fi-
ine. A fixed concentration of the substrate dextromethorphan
was incubated with liver microsomes, appropriate cofactors, and nally, data from cDNA-expressed human cytochromes can
varying concentrations of perphenazine or clozapine. Rates of be misinterpreted unless they are considered in the correct
formation of dextrorphan (mediated by CYP2D6) with inhibitor context. Pure cytochrome studies can yield quantitative data
present were expressed as a ratio versus the control velocity with
no inhibitor. The relation of the velocity ratio to inhibitor concen- on the activity of one or more particular cytochromes as a
tration can be used to calculate a 50% inhibitory concentration mediator of a specific reaction. However, this information
(IC50). The results indicate that perphenazine (IC50 0.9 M) is cannot be extrapolated to an estimate of the relative activity
likely to be a clinically important inhibitor of CYP2D6, whereas
clozapine (IC50 119 M) is a weak inhibitor. of different cytochromes, either in vivo or in liver micro-
somes in vitro, without a parallel estimate of the relative
quantitative abundance of the cytochromes in question.
That is, the importance of a specific cytochrome depends
on both activity and abundance.
can be applied to in vitroin vivo scaling only when the
mechanism of inhibition is noncompetitive (111,112). In-
hibitory potency can also be estimated by determining the DRUG INTERACTIONS IN
inhibition constant (Ki), a number that reflects inhibitory PSYCHOPHARMACOLOGY
activity in reciprocal fashion. Estimation of Ki requires the
study of multiple substrate concentrations and multiple in- During the last two decades the general problem of pharma-
hibitor concentrations and therefore involves more work, cokinetic drug interactions has received increased attention.
time, and expense. The numerical value of Ki depends on New classes of medications introduced into clinical practice
the specific biochemical mechanism of inhibition, which over this period include the selective serotonin reuptake
may be unknown. Nonetheless, the inhibitor Ki, is indepen- inhibitor (SSRI) and related mixed mechanism antidepres-
dent of substrate concentration and can be used under some sants, the azole antifungal agents, newer macrolide antimi-
defined circumstances for quantitative in vitroin vivo scal- crobial agents, and the highly active antiretroviral therapies
ing of drug interactions. In general, Ki is always less than (HAARTs) used against HIV infection and AIDS (Table
or equal to IC50; Ki will be essentially equal to IC50 if inhibi- 38.4). These and other classes of medications have had a
tion is noncompetitive, or if inhibition is competitive and major beneficial impact on the therapy on some serious and
the substrate concentration is far below the reaction Km. life-threatening illnesses, but many of the drugs have the
Both Ki and IC50 should provide similar or identical rank- secondary pharmacologic property of inducing or inhibiting
order estimates of relative inhibitory potency for a series of the human CYP enzymes, thereby raising concerns about
inhibitors of a specific reaction. When the inhibition is pharmacokinetic drug interactions during multiple drug
purely competitive, Ki values for a specific inhibitor should therapy.
theoretically be identical across different substrates metabo- One major objective of the drug development process
lized by that particular CYP. However, this principle is not is to generate data on drug interactions, so that treating
supported by experimental data, probably because of in vitro physicians have the information necessary for safe clinical
experimental artifacts, and because actual biochemical therapy involving multiple medications. However, the
mechanisms of inhibition are not purely competitive. number of possible drug interactions is very large, and time
Therefore, absolute values of Ki or IC50 cannot be assumed and resources available for implementation of controlled
to cross different substrates for the same cytochrome, al- clinical pharmacokinetic studies are inevitably limited.
518 Neuropsychopharmacology: The Fifth Generation of Progress

Some needed studies will therefore be postponed until after lished. In any case, the theoretical assumption that unbound
a new drug is marketed, and some studies may be bypassed plasma concentrations are equal to enzyme-available intra-
altogether. As discussed above, in vitro data are becoming hepatic concentrations is incorrect in reality, and may yield
increasingly important as a resource for identifying proba- underestimates of observed in vivo drug interactions by as
ble, possible, or unlikely drug interactions, and thereby en- much as an order of magnitude or more (8,83,85).
couraging rational planning and allocation of resources to Induction of CYP-mediated metabolism requires prior
more definitive clinical studies. exposure to a chemical inducer, which signals the synthetic
mechanisms to upregulate the production of one or more
CYP isoforms (114118). This process takes time, and the
Pharmacokinetic Versus increase in CYP activity is of slow onset following initiation
Pharmacodynamic Drug Interactions of exposure to the inducer, and slowly reverts to baseline
after the inducer is removed. Increased CYP expression/
A pharmacokinetic interaction implies that the drug pro-
activity due to chemical induction therefore reflects prior
ducing the interaction (the perpetrator) causes a change
but not necessarily current exposure to the inducer. The
in the metabolic clearance of the drug being affected by
extent of CYP induction probably depends on the dosage
the interaction (the victim), in turn either decreasing or
(concentration) of the inducer and on the duration of expo-
increasing concentrations of the victim drug in plasma and
sure. Induction, unlike inhibition, is not easily studied in
presumably also at the site of action. This change may or
vitro, because induction requires intact cellular protein syn-
may not alter the clinical activity of the victim drug. One
thesis mechanisms as are available in cell culture models.
pharmacokinetic interaction variant involves modification
Inducers and inhibitors of CYP3A can be expected to
by the perpetrator of the victim drugs access to its pharma-
influence both hepatic and gastrointestinal CYP3A, al-
cologic receptor site, without changing the systemic clear-
though not necessarily to the same extent. Very strong in-
ance or plasma levels of the victim. A familiar example is
hibitors (such as ketoconazole) or very strong inducers (such
the antagonism of benzodiazepine agonist activity by fluma-
as rifampin) will produce substantial changes in both he-
zenil; a less familiar example is benzodiazepine receptor an-
patic and gastrointestinal CYP3A. A uniquely complex situ-
tagonism by ketoconazole (113).
ation arises for ritonavir, which is both an inhibitor and
A pharmacodynamic interaction involves either inhibi-
inducer of CYP3A. Interactions of ritonavir with CYP3A
tion or enhancement of the clinical effects of the victim
substrate drugs will be time dependent. Initial exposure will
drug as a result of similar or identical end-organ actions.
produce CYP3A inhibition, but as the duration of exposure
Examples are the increase or decrease of the sedative-hyp-
proceeds, CYP3A induction may offset the inhibitory effects
notic actions of benzodiazepines due to coadministration
of acute exposure. The net outcome typically is unpredict-
of ethanol or caffeine, respectively.
able and variable across individuals (9395).

Mechanisms of Inhibition Versus Perspectives on the Clinical Importance

Induction of Metabolism of Drug Interactions
Drug interactions due to inhibition as opposed to induction Based on the prevalence of polypharmacy in clinical prac-
of CYP-mediated metabolism involve mechanistically dif- tice, noninteractions of drugs are far more common than
ferent processes. Chemical inhibition is an immediate phe- interactions. Coadministration of two drugs usually pro-
nomenon that becomes evident as soon as the inhibitor duces no detectable pharmacokinetic or pharmacodynamic
comes in contact with the enzyme, and is in principle revers- interaction, and the pharmacokinetic disposition and clini-
ible when the inhibitor is no longer present [an exception cal activity of each drug proceed independently of each
is mechanism-based inhibition (86)]. The magnitude of other. A less common outcome of drug coadministration is
inhibition depends on the inhibitor concentration at the a kinetic interaction that could be detected in controlled
site of the enzyme in relation to the intrinsic potency of laboratory circumstances, but that is not clinically impor-
the inhibitor. In vitro systems can be used to develop quanti- tant in usual therapeutic circumstances because (a) the inter-
tative estimates of inhibitory potency, such as the inhibition action, whether or not statistically significant, is not large
constant (Ki) or the 50% inhibitory concentration (IC50) enough to produce a clinically important change in dynam-
(Fig. 38.13). However, application of Ki or IC50 values from ics of the victim drug; (b) the therapeutic index of the victim
in vitro systems to quantitative predictions of drug interac- drug is large enough so that even a substantial change in
tions in vivo is not straightforward, and requires knowledge plasma levels will not alter therapeutic effects or toxicity;
of the effective concentration of inhibitor that is available or (c) kinetics and response to the victim drug are so variable
to the enzyme. A generally applicable scheme for relating that changes in plasma levels due to the drug interaction
total or unbound plasma concentrations of inhibitor to ef- are far less important than inherent variability. Even less
fective enzyme-available concentration has not been estab- common are clinically important interactions that require
 38: Pharmacokinetics, Pharmacodynamics, and Drug Disposition 519

modification in dosage of the perpetrator, the victim, or compounds produce substantial in vitro inhibition of triazo-
both. Finally, the most unusual consequence of a drug inter- lam hydroxylation and have the potential to produce a sig-
action is a hazardous and contraindicated combination, as nificant interaction with triazolam in vivo. However,
in the case of ketoconazole and terfenadine. These situations azithromycin was a very weak inhibitor of triazolam in vitro
are rare, but unfortunately receive excessive attention in the (IC50 250 M), and is anticipated to produce no signifi-
public media. cant interaction in vivo.
Many secondary sources and compendia are available as In a clinical pharmacokinetic-pharmacodynamic study
summary guides to the extensive literature on drug interac- (123), a series of healthy volunteers were exposed to the
tions, but these sources do not necessarily assist clinicians following treatment conditions:
in deciding which interactions should generate serious con-
A. Triazolam placebo plus macrolide placebo
cern in the course of drug therapy. A useful general guideline
B. Triazolam (0.125 mg) plus macrolide placebo
is that drug interactions are more likely to be important
C. Triazolam (0.125 mg) plus azithromycin
when (a) the perpetrator drug is a powerful inducer or inhib-
D. Triazolam (0.125 mg) plus erythromycin
itor, and produces a very large change in the kinetics and
E. Triazolam (0.125 mg) plus clarithromycin
plasma levels of the victim drug; or (b) the therapeutic index
of the victim is narrow. Case (a) is exemplified by powerful Dosage schedules of the coadministered macrolides were
inducers or inhibitors of CYP3A (ketoconazole, ritonavir, chosen to be consistent with usual dosage recommenda-
rifampin) coadministered with CYP3A substrates, or power- tions. The five trials were randomized in sequence, and the
ful inhibitors of CYP2D6 (quinidine, fluoxetine, parox- treatment conditions were double-blind.
etine) coadministered with CYP2D6 substrates. Case (b) is Following each dose of triazolam (or placebo to match
exemplified by victim drugs such as phenytoin, warfarin, triazolam), multiple venous blood samples were drawn over
and digoxin, for which small changes in plasma levels could a period of 24 hours, and multiple pharmacodynamic test-
have important clinical consequences. ing procedures were performed. Triazolam plasma concen-
trations were determined by gas chromatography with elec-
tron capture detection (Fig. 38.14).
Application of Kinetic-Dynamic Methods
to Study Drug Interactions
Drug interaction study protocols often incorporate pharma-
codynamic endpoints to allow estimating the clinical conse-
quences of drug interactions along with the usual pharmaco-
kinetic outcome measures. The level of complexity of an
integrated kinetic-dynamic study depends on the nature of
the pharmacodynamic actions of the drug under study as
well as the type of pharmacodynamic outcome measures
that are required. A number of methodologic principles and
dilemmas are illustrated by kinetic-dynamic design options
for drug interaction studies involving sedative-hypnotic and
anxiolytic drugs acting on the -aminobutyric acid
(GABA)-benzodiazepine receptor system.
Biotransformation of the benzodiazepine triazolam is de-
pendent on the activity of human CYP3A isoforms (119).
Metabolism is strongly inhibited in vitro and in vivo by
CYP3A inhibitors such as ketoconazole, itraconazole, rito-
navir, and nefazodone (95,119122). Some, but not all, of
the macrolide antimicrobial agents also are CYP3A inhibi-
tors via mechanism-based inhibition, in which the par-
ent compound binds to the metabolically active site on
FIGURE 38.14. Mean plasma triazolam concentrations following
the CYP3A enzyme, yielding a metabolic intermediate single 0.125-mg doses of triazolam during trials B, C, D, and E.
that inactivates the enzyme (86,87). We tested the inhibi- Note that coadministration of triazolam with azithromycin (AZI,
tory potency of four macrolide antimicrobial agents trial C) produced plasma levels nearly identical to triazolam ad-
ministered with placebo (PL, trial B). However, coadministration
[troleandomycin (TAO), erythromycin, clarithromycin, with erythromycin (ERY, trial D) or clarithromycin (CLAR, trial E)
azithromycin] versus triazolam hydroxylation using human produced a large increase in plasma triazolam concentrations.
liver microsomes in vitro (123). Appropriate mean IC50 val- (Adapted in part from Greenblatt DJ, von Moltke LL, Harmatz JS,
et al. Inhibition of triazolam clearance by macrolide antimicrobial
ues were TAO, 3.6 M; erythromycin, 30 M; and clar- agents: in vitro correlates and dynamic consequences. Clin Phar-
ithromycin, 28 M. These values indicate that all three macol Ther 1998;64:278285, with permission.)
520 Neuropsychopharmacology: The Fifth Generation of Progress

FIGURE 38.15. Mean ( standard

error, SE) 4-hour pharmacodynamic
effect areas for the digit-symbol sub-
stitution test (DSST) score (left), and
for the EEG beta amplitude (right),
during the five trials. Note that decre-
ments in DSST score, and increases in
EEG beta amplitude, were very similar
between trials B and C, whereas ef-
fects were significantly enhanced
during trials D and E.

Mean clearance of triazolam during trials B and C was macokinetics, pharmacodynamics, and drug metabolism
nearly identical (413 and 416 mL/min, respectively); that may be usefully applied to the evaluation of drug interac-
is, coadministration of azithromycin had no effect on the tions. An ideal approach would incorporate the collabora-
pharmacokinetics of triazolam (Fig. 38.14). However, tria- tive participation of individuals representing expertise in
zolam clearance was significantly reduced to 146 mL/min molecular pharmacology, cytochrome biochemistry, in vitro
by erythromycin (trial D), and to 95 mL/min by clarithro- metabolism, clinical pharmacokinetics-pharmacodynamics,
mycin (trial E) (Fig. 38.14). Thus the in vivo kinetic results and clinical therapeutics.
are highly consistent with the in vitro data.
The pharmacodynamic data indicated that the benzodi-
azepine agonist effects of triazolam plus placebo (trial B),
and of triazolam plus azithromycin (trial C) were similar to
each other, and greater than the effects of placebo plus pla-
cebo (trial A). However, coadministration of triazolam with
erythromycin (trial D) or with clarithromycin (trial E) aug-
mented the pharmacodynamic effects of triazolam when
compared to trials B or C. The outcome was similar whether
based on subjective measures, a semi-objective measure (the
Digit-Symbol Substitution Test, DSST), or the fully objec-
tive measure (the EEG) (Fig. 38.15). Kinetic-dynamic mod-
eling indicated that the increase in benzodiazepine agonist
effects of triazolam caused by coadministration of erythro-
mycin or clarithromycin was fully consistent with the in-
crease in triazolam plasma concentrations (Fig. 38.16).

COMMENT

Pharmacokinetic drug interactions in clinical psychophar-

macology are assuming increasing importance as polyphar-
macy becomes more common, and more drugs with en- FIGURE 38.16. Relation of mean plasma triazolam concentra-
tions to mean changes over baseline in DSST score at the corre-
zyme-inducing or -inhibiting properties are introduced into sponding times. The solid line represents the kinetic-dynamic
clinical practice. Contemporary approaches to the basic and model relationship based on an exponential function as shown
clinical investigation of drug interactions and their pharma- in Fig. 38.5. (From Greenblatt DJ, von Moltke LL, Harmatz JS, et
al. Inhibition of triazolam clearance by macrolide antimicrobial
codynamic consequences are illustrated in this chapter. It agents: in vitro correlates and dynamic consequences. Clin Phar-
is evident that technologic and conceptual advances in phar- macol Ther 1998;64:278285, with permission.)
 38: Pharmacokinetics, Pharmacodynamics, and Drug Disposition 521

ACKNOWLEDGMENTS 18. Aarons L. Population pharmacokinetics: theory and practice.

Br J Clin Pharmacol 1991;32:669670.
19. de Gatta MMF, Garca MJ, Lanao JM, et al. Bayesian forecast-
The work was supported by grants MH-34223, MH- ing in paediatric populations. Clin Pharmacokinet 1996;31:
01237, DA-05258, DA-13209, MH-58435, DK-58496, 325330.
and RR-00054 from the U.S. Department of Health and 20. Shader RI, Harmatz JS, Oesterheld JR, et al. Population phar-
Human Services. macokinetics of methylphenidate in children with attention-
deficit hyperactivity disorder. J Clin Pharmacol 1999;39:
775785.
21. Safer DJ. Central stimulant treatment of childhood attention
deficit hyperactivity disorder. CNS Drugs 1997;7:264272.
REFERENCES 22. Elia J, Ambrosini PJ, Rapoport JL. Treatment of attention-
1. Stanski DR. Pharmacodynamic modeling of anesthetic EEG deficit-hyperactivity disorder. N Engl J Med 1999;340:
drug effects. Annu Rev Pharmacol Toxicol 1992;32:423447. 780788.
2. Derendorf H, Meibohm B. Modeling of pharmacokinetic/phar- 23. Cyr M, Brown CS. Current drug therapy recommendations for
macodynamic (PK/PD) relationships: concepts and perspec- the treatment of attention deficit hyperactivity disorder. Drugs
tives. Pharm Res 1999;16:176185. 1998;56:215233.
3. Bellissant E, Sebille V, Paintaud G. Methodological issues in 24. Wilens TE, Biederman J, Spencer TJ. Pharmacotherapy of at-
pharmacokinetic-pharmacodynamic modelling. Clin Pharma- tention deficit hyperactivity disorder in adults. CNS Drugs
cokinet 1998;35:151166. 1998;9:347356.
4. Mandema JW, Danhof M. Electroencephalogram effect mea- 25. Kimko HC, Cross JT, Abernethy DR. Pharmacokinetics and
sures and relationships between pharmacokinetics and pharma- clinical effectiveness of methylphenidate. Clin Pharmacokinet
codynamics of centrally acting drugs. Clin Pharmacokinet 1992; 1999;37:457470.
23:191215. 26. Srinivas NR, Hubbard JW, Quinn D, et al. Enantioselective
5. Laurijssens BE, Greenblatt DJ. Pharmacokinetic-pharmacody- pharmacokinetics and pharmacodynamics of dl-threo-methyl-
namic relationships for benzodiazepines. Clin Pharmacokinet phenidate in children with attention deficit hyperactivity disor-
1996;30:5276. der. Clin Pharmacol Ther 1992;52:561568.
6. Greenblatt DJ, von Moltke LL, Harmatz JS, et al. Human cyto- 27. Schwinghammer TL, Kroboth PD. Basic concepts in pharmaco-
chromes and some newer antidepressants: kinetics, metabolism, dynamic modeling. J Clin Pharmacol 1988;28:388394.
and drug interactions. J Clin Psychopharmacol 1999;19(suppl 28. Holford NH, Sheiner LB. Understanding the dose-effect rela-
1):23S35S. tionship: clinical application of pharmacokinetic-pharmacody-
7. Greenblatt DJ, von Moltke LL, Harmatz JS, et al. Drug interac- namic models. Clin Pharmacokinet 1981;6:429453.
tions with newer antidepressants: role of human cytochromes 29. Swerdlow BN, Holley FO. Intravenous anaesthetic agents.
P450. J Clin Psychiatry 1998;59(suppl 15):1927. Pharmacokinetic-pharmacodynamic relationships. Clin Phar-
8. von Moltke LL, Greenblatt DJ, Schmider J, et al. In vitro ap- macokinet 1987;12:79110.
proaches to predicting drug interactions in vivo. Biochem Phar- 30. Dingemanse J, Danhof M, Breimer DD. Pharmacokinetic-
macol 1998;55:113122. pharmacodynamic modeling of CNS drug effects: an overview.
9. Bertz RJ, Granneman GR. Use of in vitro and in vivo data to Pharmacol Ther 1988;38:152.
estimate the likelihood of metabolic pharmacokinetic interac- 31. Campbell DB. The use of kinetic-dynamic interactions in the
tions. Clin Pharmacokinet 1997;32:210258. evaluation of drugs. Psychopharmacology 1990;100:433450.
10. Whiting B, Kelman AW, Grevel J. Population pharmacoki- 32. Greenblatt DJ, Harmatz JS. Kinetic-dynamic modeling in clini-
netics. Theory and clinical application. Clin Pharmacokinet cal psychopharmacology. J Clin Psychopharmacol 1993;13:
1986;11:387401. 231234.
11. Vozeh S, Steimer JL, Rowland M, et al. The use of population 33. Greenblatt DJ, von Moltke LL, Ehrenberg BL, et al. Kinetics
pharmacokinetics in drug development. Clin Pharmacokinet and dynamics of lorazepam during and after continuous intrave-
1996;30:8193. nous infusion. Crit Care Med 2000;28:27502757.
12. Maitre PO, Buhrer M, Thomson D, et al. A three-step approach 34. Greenblatt DJ, Ehrenberg BL, Gunderman J, et al. Kinetic and
combining Bayesian regression and NONMEM population dynamic study of intravenous lorazepam: comparison with in-
analysis: application to midazolam. J Pharmacokinet Biopharm travenous diazepam. J Pharmacol Exp Ther 1989;250:134140.
1991;19:377384. 35. Greenblatt DJ, Ehrenberg BL, Gunderman J, et al. Pharmacoki-
13. Thomson AH, Whiting B. Bayesian parameter estimation and netic and electroencephalographic study of intravenous diaze-
population pharmacokinetics. Clin Pharmacokinet 1992;22: pam, midazolam, and placebo. Clin Pharmacol Ther 1989;45:
447467. 356365.
14. Jonsson EN, Wade JR, Karlsson MO. Comparison of some 36. Gupta SK, Ellinwood EH, Nikaido AM, et al. Simultaneous
practical sampling strategies for population pharmacokinetic modeling of the pharmacokinetic and pharmacodynamic prop-
studies. J Pharmacokinet Biopharm 1996;24:245263. erties of benzodiazepines. I: Lorazepam. J Pharmacokinet Bio-
15. Carter AA, Rosenbaum SE, Dudley MN. Review of methods pharm 1990;18:89102.
in population pharmacokinetics. Clin Res Regul Affairs 1995; 37. Tedeschi G, Smith AT, Dhillon S, et al. Rate of entrance of
12:121. benzodiazepines into the brain determined by eye movement
16. Samara E, Granneman R. Role of population pharmacokinetics recording. Br J Clin Pharmacol 1983;15:103107.
in drug development: a pharmaceutical industry perspective. 38. Greenblatt DJ, Sethy VH. Benzodiazepine concentrations in
Clin Pharmacokinet 1997;32:294312. brain directly reflect receptor occupancy: studies of diazepam,
17. Sheiner LB, Ludden TM. Population pharmacokinetics/dynam- lorazepam, and oxazepam. Psychopharmacology 1990;102:
ics. Annu Rev Pharmacol Toxicol 1992;32:185209. 373378.
522 Neuropsychopharmacology: The Fifth Generation of Progress

39. Walton NY, Treiman DM. Lorazepam treatment of experimen- 61. Fromm MF, Kroemer HK, Eichelbaum M. Impact of P450
tal status epilepticus in the rat: relevance to clinical practice. genetic polymorphism on the first-pass extraction of cardiovas-
Neurology 1990;40:990994. cular and neuroactive drugs. Advanced Drug Deliv Rev 1997;
40. Ameer B, Greenblatt DJ. Lorazepam: a review of its clinical 27:171199.
pharmacological properties and therapeutic uses. Drugs 1981; 62. Gonzalez FJ, Meyer UA. Molecular genetics of the debrisoquin-
21:161200. sparteine polymorphism. Clin Pharmacol Ther 1991;50:
41. Greenblatt DJ. Clinical pharmacokinetics of oxazepam and lor- 233238.
azepam. Clin Pharmacokinet 1981;6:88105. 63. Guengerich FP. Cytochrome P-450 3A4: regulation and role
42. Ochs HR, Greenblatt DJ, Knuchel M. Kinetics of diazepam, in drug metabolism. Annu Rev Pharmacol Toxicol 1999;39:1
midazolam, and lorazepam in cigarette smokers. Chest 1985; 7.
87:223226. 64. Dresser GK, Spence JD, Bailey DG. Pharmacokinetic-pharma-
43. Abernethy DR, Greenblatt DJ, Ameer B, et al. Probenecid im- codynamic consequences and clinical relevance of cytochrome
pairment of acetaminophen and lorazepam clearance: direct in- P450 3A4 inhibition. Clin Pharmacokinet 2000;38:4157.
hibition of ether glucuronide formation. J Pharmacol Exp Ther 65. Thummel KE, Wilkinson GR. In vitro and in vivo drug interac-
1985;234:345349. tions involving human CYP3A. Annu Rev Pharmacol Toxicol
44. Abernethy DR, Greenblatt DJ, Divoll M, et al. Differential 1998;38:389430.
effect of cimetidine on drug oxidation (antipyrine and diaze- 66. de Wildt SN, Kearns GL, Leeder JS, et al. Cytochrome P450
pam) versus conjugation (acetaminophen and lorazepam): pre- 3A: ontogeny and drug disposition. Clin Pharmacokinet 1999;
vention of acetaminophen toxicity by cimetidine. J Pharmacol 37:485505.
Exp Ther 1983;224:508513. 67. Ketter TA, Flockhart DA, Post RM, et al. The emerging role
45. Greenblatt DJ, Allen MD, Locniskar A, et al. Lorazepam kinet- of cytochrome P450 3A in psychopharmacology. J Clin Psycho-
ics in the elderly. Clin Pharmacol Ther 1979;26:103113. pharmacol 1995;15:387398.
46. Treiman DM. The role of benzodiazepines in the management 68. Maurel P. The CYP3A family. In: Ionnides C, ed. Cytochromes
of status epilepticus. Neurology 1990;40(suppl 2):3242. P450. Boca Raton, FL: CRC Press, 1996:241270.
47. Lowenstein DH, Alldredge BK. Status epilepticus. N Engl J 69. von Moltke LL, Greenblatt DJ, Schmider J, et al. Metabolism
Med 1998;338:970976. of drugs by cytochrome P450 3A isoforms: implications for
48. Clarke SE. In vitro assessment of human cytochrome P450. drug interactions in psychopharmacology. Clin Pharmacokinet
Xenobiotica 1998;28:11671202. 1995;29(suppl 1):3343.
49. Smith G, Stubbins MJ, Harries LW, et al. Molecular genetics 70. Shimada T, Yamazaki H, Mimura M, et al. Interindividual
of the human cytochrome P450 monooxygenase superfamily. variations in human liver cytochrome P-450 enzymes involved
Xenobiotica 1998;28:11291165. in the oxidation of drugs, carcinogens and toxic chemicals: stud-
50. Smith DA, Abel SM, Hyland R, et al. Human cytochrome ies with liver microsomes of 30 Japanese and 30 Caucasians. J
P450s: selectivity and measurement in vivo. Xenobiotica 1998; Pharmacol Exp Ther 1994;270:414423.
28:10951128. 71. Venkatakrishnan K, Greenblatt DJ, von Moltke LL, et al. Five
51. Nelson DR, Koymans L, Kamataki T, et al. P450 superfamily: distinct human cytochromes mediate amitriptyline N-demeth-
update on new sequences, gene mapping, accession numbers ylation in vitro: dominance of CYP 2C19 and 3A4. J Clin
and nomenclature. Pharmacogenetics 1996;6:142. Pharmacol 1998;38:112121.
52. Park BK, Pirmohamed M, Kitteringham NR. The role of cyto- 72. von Moltke LL, Greenblatt DJ, Grassi JM, et al. Citalopram and
chrome P450 enzymes in hepatic and extrahepatic human drug desmethylcitalopram in vitro: human cytochromes mediating
toxicity. Pharmacol Ther 1995;68:385424. transformation, and cytochrome inhibitory effects. Biol Psychia-
53. Wrighton SA, Stevens JC. The human hepatic cytochromes try 1999;46:839849.
P450 involved in drug metabolism. Crit Rev Toxicol 1992;22: 73. von Moltke LL, Greenblatt DJ, Granda BW, et al. Zolpidem
121. metabolism in vitro: responsible cytochromes, chemical inhibi-
54. Glue P, Clement RP. Cytochrome P450 enzymes and drug tors, and in vivo correlations. Br J Clin Pharmacol 1999;48:
metabolismbasic concepts and methods of assessment. Cell 8997.
Mol Neurobiol 1999;19:309323. 74. von Moltke LL, Greenblatt DJ, Granda BW, et al. Nefazodone,
55. Parkinson A. An overview of current cytochrome P450 technol- meta-chlorophenylpiperazine, and their metabolites in vitro: cy-
ogy for assessing the safety and efficacy of new materials. Toxicol tochromes mediating transformation, and P4503A4 inhibitory
Pathol 1996;24:4557. actions. Psychopharmacology 1999;145:113122.
56. Kroemer HK, Eichelbaum M. Molecular bases and clinical con- 75. Greene DS, Barbhaiya RH. Clinical pharmacokinetics of nefa-
sequences of genetic cytochrome P450 2D6 polymorphism. Life zodone. Clin Pharmacokinet 1997;33:260275.
Sci 1995;56:22852298. 76. von Moltke LL, Greenblatt DJ, Schmider J, et al. Midazolam
57. Nebert DW. Polymorphisms in drug-metabolizing enzymes: hydroxylation by human liver microsomes in vitro: inhibition
what is their clinical relevance and why do they exist? Am J by fluoxetine, norfluoxetine, and by azole antifungal agents. J
Hum Genet 1997;60:265271. Clin Pharmacol 1996;36:783791.
58. Ingelman-Sundberg M, Oscarson M, McLellan RA. Polymor- 77. Perloff MD, von Moltke LL, Court MH, et al. Midazolam and
phic human cytochrome P450 enzymes: an opportunity for in- triazolam biotransformation in mouse and human liver micro-
dividualized drug treatment. Trends Pharmacol Sci 1999;20: somes: relative contribution of CYP3A and CYP2C9 isoforms.
342349. J Pharmacol Exp Ther 2000;292:618628.
59. Bertilsson L. Geographical/interracial differences in polymor- 78. Bornemann LD, Min BH, Crews T, et al. Dose dependent
phic drug oxidation: current state of knowledge of cytochromes pharmacokinetics of midazolam. Eur J Clin Pharmacol 1985;
P450 (CYP) 2D6 and 2C19. Clin Pharmacokinet 1995;29: 29:9195.
192209. 79. Hall SD, Thummel KE, Watkins PB, et al. Molecular and physi-
60. Bertilsson L, Dahl M-L. Polymorphic drug oxidation. CNS cal mechanisms of first-pass extraction. Drug Metab Dispos
Drugs 1996;3:200223. 1999;27:161166.
 38: Pharmacokinetics, Pharmacodynamics, and Drug Disposition 523

80. Yuan R, Flockhart DA, Balian JD. Pharmacokinetic and phar- 101. Ball SE, Scatina A, Kao J, et al. Population distribution and
macodynamic consequences of metabolism-based drug interac- effects on drug metabolism of a genetic variant in the 5 pro-
tions with alprazolam, midazolam, and triazolam. J Clin Phar- moter region of CYP3A4. Clin Pharmacol Ther 1999;66:
macol 1999;39:11091125. 288294.
81. Tsunoda SM, Velez RL, von Moltke LL, et al. Differentiation 102. Sata F, Sapone A, Elizondo G, Stocker P, et al. CYP3A4 allelic
of intestinal and hepatic cytochrome P450 3A activity with use variants with amino acid substitutions in exons 7 and 12: Evi-
of midazolam as an in vivo probe: effect of ketoconazole. Clin dence for an allelic variant with altered catalytic activity. Clin
Pharmacol Ther 1999;66:461471. Pharmacol Ther 2000;67:4856.
82. Lin JH, Lu AY. Inhibition and induction of cytochrome P450 103. von Moltke LL, Tran TH, Cotreau MM, et al. Unusually low
and the clinical implications. Clin Pharmacokinet 1998;35: clearance of two CYP3A4 substrates, alprazolam and trazodone,
361390. in a volunteer subject with wild-type CYP3A promoter region.
83. Venkatakrishnan K, von Moltke LL, Greenblatt DJ. Effects of J Clin Pharmacol 2000;40:200204.
the antifungal agents on oxidative drug metabolism in humans: 104. Westlind A, Lofberg L, Tindberg N, et al. Interindividual differ-
clinical relevance. Clin Pharmacokinet 2000;38:111180. ences in hepatic expression of CYP3A4: relationship to genetic
84. Ito K, Iwatsubo T, Kanamitsu S, et al. Prediction of pharmaco- polymorphism in the 5-upstream regulatory region. Biochem
kinetic alterations caused by drug-drug interactions: metabolic Biophys Res Commun 1999;259:201205.
interaction in the liver. Pharmacol Rev 1998;50:387412. 105. Wandel C, Witte JS, Hall JM, et al. CYP3A activity in African
85. Greenblatt DJ, von Moltke LL. Sedative-hypnotic and anxio- American and European American men: population differences
lytic agents. In: Levy RH, Thummel KE, Trager WF, et al., and functional effect of CYP3A4*1B 5-promoter region poly-
eds. Metabolic drug interactions. Philadelphia: Lippincott Wil- morphism. Clin Pharmacol Ther 2000;68:8291.
liams & Wilkins, 2000:259270. 106. Gelboin HV, Krausz KW, Gonzalez FJ, et al. Inhibitory mono-
86. Silverman R. Mechanism-based enzyme inactivators. Methods clonal antibodies to human cytochrome P450 enzymes: a new
Enzymol 1995;249:240283. avenue for drug discovery. Trends Pharmacol Sci 1999;20:
87. Gillum JG, Israel DS, Polk RE. Pharmacokinetic drug interac-
432438.
tions with antimicrobial agents. Clin Pharmacokinet 1993;25:
107. Shou M, Lu T, Krausz KW, et al. Use of inhibitory monoclonal
450482.
antibodies to assess the contribution of cytochromes P450 to
88. Barry M, Mulcahy F, Merry C, et al. Pharmacokinetics and
human drug metabolism. Eur J Pharmacol 2000;394:199
potential interactions amongst antiretroviral agents used to treat
209.
patients with HIV infection. Clin Pharmacokinet 1999;36:
108. Gonzalez FJ, Korzekwa KR. Cytochromes P450 expression sys-
289304.
tems. Annu Rev Pharmacol Toxicol 1995;35:369390.
89. Malaty LI, Kuper JJ. Drug interactions of HIV protease inhibi-
tors. Drug Safety 1999;20:147169. 109. Crespi CL, Miller VP. The use of heterologously expressed drug
90. Tseng AL, Foisy MM. Significant interactions with new antiret- metabolizing enzymesstate of the art and prospects for the
rovirals and psychotropic drugs. Ann Pharmacother 1999;33: future. Pharmacol Ther 1999;84:121131.
461473. 110. Crespi CL, Penman BW. Use of cDNA-expressed human cyto-
91. Hsu A, Granneman GR, Bertz RJ. Ritonavir. Clinical pharma- chrome P450 enzymes to study potential drug-drug interac-
cokinetics and interactions with other anti-HIV agents. Clin tions. Adv Pharmacol 1997;43:171188.
Pharmacokinet 1998;35:275291. 111. Halpert JR. Structural basis of selective cytochrome P450 inhi-
92. Bailey DG, Malcom J, Arnold O, et al. Grapefruit juicedrug bition. Annu Rev Pharmacol Toxicol 1995;35:2953.
interactions. Br J Clin Pharmacol 1998;46:101110. 112. Segel IH. Enzyme kinetics. New York: Wiley, 1975.
93. Greenblatt DJ, von Moltke LL, Harmatz JS, et al. Alprazolam- 113. Fahey JM, Pritchard GA, von Moltke LL, et al. The effects of
ritonavir interaction: implications for product labeling. Clin ketoconazole on triazolam pharmacokinetics, pharmacodynam-
Pharmacol Ther 2000;67:335341. ics and benzodiazepine receptor binding in mice. J Pharmacol
94. Greenblatt DJ, von Moltke LL, Daily JP, et al. Extensive impair- Exp Ther 1998;285:271276.
ment of triazolam and alprazolam clearance by short-term low- 114. Barry M, Feely J. Enzyme induction and inhibition. Pharmacol
dose ritonavir: the clinical dilemma of concurrent inhibition Ther 1990;48:7194.
and induction. J Clin Psychopharmacol 1999;19:293296. 115. Denison MS, Whitlock JP. Xenobiotic-inducible transcription
95. Greenblatt DJ, von Moltke LL, Harmatz JS, et al. Differential of cytochromes P450 genes. J Biol Chem 1995;270:
impairment of triazolam and zolpidem clearance by ritonavir. 1817518178.
J AIDS 2000;24:129136. 116. Bock KW, Lipp H-P, Bock-Hennig BS. Induction of drug-
96. Kassai A, Toth G, Eichelbaum M, et al. No evidence of a genetic metabolizing enzymes by xenobiotics. Xenobiotica 1990;20:
polymorphism in the oxidative metabolism of midazolam. Clin 11011111.
Pharmacokinet 1988;15:319325. 117. Waxman DJ, Azaroff L. Phenobarbital induction of cytochrome
97. Friedman H, Greenblatt DJ, Burstein ES, et al. Population P-450 gene expression. Biochem J 1992;281:577592.
study of triazolam pharmacokinetics. Br J Clin Pharmacol 1986; 118. Park BK, Kitteringham NR, Piromohamed M, et al. Relevance
22:639642. of induction of human drug-metabolizing enzymes: pharmaco-
98. Greenblatt DJ, Divoll M, Abernethy DR, et al. Reduced clear- logical and toxicological implications. Br J Clin Pharmacol 1996;
ance of triazolam in old age: relation to antipyrine oxidizing 41:477491.
capacity. Br J Clin Pharmacol 1983;15:303309. 119. von Moltke LL, Greenblatt DJ, Harmatz JS, et al. Triazolam
99. Felix CA, Walker AH, Lange BJ, et al. Association of CYP3A4 biotransformation by human liver microsomes in vitro: effects
genotype with treatment-related leukemia. Proc Natl Acad Sci of metabolic inhibitors, and clinical confirmation of a predicted
USA 1998;95:1317613181. interaction with ketoconazole. J Pharmacol Exp Ther 1996;276:
100. Rebbeck TR, Jaffe JM, Walker AH, et al. Modification of clini- 370379.
cal presentation of prostate tumors by a novel genetic variant 120. von Moltke LL, Greenblatt DJ, Duan SX, et al. Inhibition of
in CYP3A4. J Natl Cancer Inst 1998;90:12251229. triazolam hydroxylation by ketoconazole, itraconazole, hydroxy-
524 Neuropsychopharmacology: The Fifth Generation of Progress

itraconazole and fluconazole in vitro. Pharm Pharmacol Com- kinetic and dynamic consequences. Clin Pharmacol Ther 1998;
mun 1998;4:443445. 64:237247.
121. von Moltke LL, Greenblatt DJ, Grassi JM, et al. Protease inhibi- 123. Greenblatt DJ, von Moltke LL, Harmatz JS, et al. Inhibition
tors as inhibitors of human cytochromes P450: high risk associ- of triazolam clearance by macrolide antimicrobial agents: in
ated with ritonavir. J Clin Pharmacol 1998;38:106111. vitro correlates and dynamic consequences. Clin Pharmacol Ther
122. Greenblatt DJ, Wright CE, von Moltke LL, et al. Ketoconazole 1998;64:278285.
inhibition of triazolam and alprazolam clearance: differential

Neuropsychopharmacology: The Fifth Generation of Progress. Edited by Kenneth L. Davis, Dennis Charney, Joseph T. Coyle, and
Charles Nemeroff. American College of Neuropsychopharmacology 2002.