030-38markowitz119815.

qxd 1/18/02 2:51 PM Page 30

The effects of probenecid on the

disposition of risperidone and olanzapine
in healthy volunteers
Study Design: The metabolic pathways of most xenobiotics and endogenous compounds can be divided
into phase 1 (oxidative, reductive, and hydrolytic) and phase 2 (glucuronidation, sulfate conjugation,
glycine and glutathione conjugation, and acetylation and methylation) processes. Oxidative metabolism
by the cytochrome P450 system has been intensively investigated compared with glucuronidation and
other conjugation pathways. The primary aim of this study was to evaluate the disposition of olanzapine
or risperidone in healthy volunteers with and without coadministration of the uridine diphosphoglu-
curonate–glucuronosyltransferase inhibitor probenecid. We hypothesized that olanzapine disposition would
be altered as a result of decreased glucuronidation, whereas risperidone disposition would be relatively
unaffected.
Methods: Our objective was to investigate whether this interaction would occur in 12 healthy volunteers,
aged 22 to 42 years, who participated in a single-dose, randomized, 4-period, double-blind, crossover
study receiving a single dose of either 5 mg olanzapine or 1 mg risperidone with and without 500 mg
probenecid (8 doses over 4 days). Multiple blood samples were analyzed by means of liquid chromatogra-
phy–tandem mass spectrometry or HPLC to assess the 48-hour time course of risperidone and olanzap-
ine. Urine was assayed for free and glucuronidated drugs.
Results: When olanzapine was administered with probenecid, statistically significant differences were
observed between plasma pharmacokinetic parameters compared with olanzapine administered alone (max-
imum concentration, P < .05; area under the plasma concentration–time curve from time zero to 24 hours,
P < .01). Clearance was not significantly different between the treatment phases. Risperidone pharmaco-
kinetic parameters were not significantly different (all parameters, P > .05).
Conclusion: Inhibition of uridine diphosphoglucuronate–glucuronosyltransferase appeared to influence
the disposition of olanzapine but not risperidone. Phase 2 metabolism may significantly influence the dis-
position of antipsychotic drugs and may be an important aspect of the variability in metabolism, partici-
pation in drug-drug interactions, and clinical response to some antipsychotic agents. (Clin Pharmacol
Ther 2002;71:30-8.)

John S. Markowitz, PharmD, C. Lindsay DeVane, PharmD, Heidi L. Liston, PharmD,

David W. Boulton, PhD, and S. Craig Risch, MD Charleston, SC

From the Departments of Pharmaceutical Sciences and Psychiatry
and Behavioral Sciences, Laboratory of Drug Disposition and
Pharmacogenetics, Medical University of South Carolina.
Received for publication April 24, 2001; accepted Aug 27, 2001.
Reprint requests: John S. Markowitz, PharmD, Institute of Psychia-
try, RM 246N, Medical University of South Carolina, 67 President
St, PO Box 250861, Charleston, SC 29425.
E-mail: markowij@musc.edu
Copyright © 2002 by the American Society for Clinical Pharmacol-
ogy and Therapeutics.
0009-9236/2002/$35.00 + 0 13/1/119815
doi:10.1067/mcp.2002.119815
Risperidone and olanzapine are the two most com-
monly used atypical antipsychotic medications in the
United States for the management of schizophrenia and
other psychotic disorders. Both medications have been
shown to have a markedly better side-effect profile than
do older conventional agents such as haloperidol and
chlorpromazine and offer improved efficacy for nega-
tive or deficit-type symptoms.1 Even with these
improvements, not all patients respond adequately to
these medications. Plasma concentrations of antipsy-

30
030-38markowitz119815.qxd 1/18/02 2:51 PM Page 31
CLINICAL PHARMACOLOGY & THERAPEUTICS
VOLUME 71, NUMBER 1
chotics have been correlated with response with both
conventional2 and some atypical agents.3,4 However,
considerable interindividual variation in disposition
exists among patients.2 The reasons for this variability
probably involve both genetic and environmental vari-
ables, but the precise causes have not been identified.
The potential exists to perturb the disposition of
the atypical antipsychotics through metabolic inhibi-
tion or induction because they are all highly metabo-
lized drugs.5 Nearly all of the published data regard-
ing the potential for antipsychotic drug-drug interac-
tions have focused on phase 1 oxidative processes
mediated by the cytochrome P450 system. 6-10
Whereas glucuronidation has been extensively stud-
ied in vitro with at least 33 families of uridine diphos-
phoglucuronate–glucuronosyltransferase (UDPGT)
identified,11 the role of phase 2 conjugative processes
such as glucuronidation in the metabolism of antipsy-
chotic agents has undergone little in vivo study.
Risperidone appears to be subject to little or no
metabolism by phase 2 metabolic processes. In con-
trast, olanzapine undergoes extensive glucuronidation,
as much as 44% of circulating plasma species being
identified as its 10-N-glucuronide.12 Other olanzapine
glucuronides also have been identified.12 Plasma con-
centrations of both olanzapine and risperidone are
known to be significantly affected by drug interactions
involving the cytochrome P450 system.13,14
The primary aim of this study was to determine
whether the disposition of olanzapine or risperidone
administered to healthy volunteers is affected by inhi-
bition of glucuronidation. Probenecid is a known
UDPGT inhibitor,15 and its coadministration allows a
test of the hypothesis that risperidone should be rela-
tively unaffected whereas olanzapine disposition should
be altered by glucuronidation inhibition. Such a find-
ing would identify an important source of potential
variability in the pharmacologic response to these
antipsychotics because they are commonly coadminis-
tered with other UDPGT substrates or inhibitors, such
as oxazepam and valproic acid, respectively, in clinical
practice.16
METHODS
Subjects
The protocol was approved by the human investiga-
tional review board of the Medical University of South
Carolina. Fourteen healthy volunteers (11 men and 3
women) 22 to 42 years old (mean, 28 years) partici-
pated after giving written informed consent. All sub-
jects were nonsmokers within 20% of ideal body weight
and were in good general health as determined by med-
Markowitz et al 31
ical history, physical examination, laboratory testing,
and 12-lead electrocardiogram. Exclusion criteria
included the use of any prescription or nonprescription
medication 1 week before initiation of the study. The
use of caffeine-containing (coffee, certain colas) and
alcoholic beverages was prohibited during the study
period.
Experimental design
Drug administration. A single-dose, randomized,
4-period, double-blind, crossover design was used for
the study. Subjects received each of 4 treatment regi-
mens in random order, each treatment period separated
by a minimum 14-day washout. During the initial phase
of the study, subjects received a single oral dose of
either 5 mg olanzapine or 1 mg risperidone contained
within identical gelatin capsules. After a minimum
14-day washout period, subjects repeated dosing
with the assigned medication but received 500 mg
probenecid, 1 tablet twice daily, 1 day before dosing of
the antipsychotic. Probenecid treatment continued the
day of antipsychotic dosing and 2 days thereafter while
blood sampling was conducted. Subjects then crossed
over to the other antipsychotic with and without coad-
ministration of probenecid.
Safety and tolerability. Baseline and periodic moni-
toring of vital signs was performed after administration
of the antipsychotic. Baseline and periodic self-assess-
ment forms were filled out by each subject to record
any side effects during the active study period. Any con-
dition present during or after the treatment period not
present at the baseline or screening examination was
documented as an adverse event and followed to a sat-
isfactory conclusion.
Analytic methods
Blood sampling. All subjects fasted for 8 hours
before drug administration and then were given a stan-
dard breakfast consisting of a bagel, cream cheese, and
orange juice. An indwelling venous catheter was placed
into an antecubital vein for serial blood sampling. After
baseline blood pressure and heart rate readings were
obtained, subjects voided the bladder and were given 5
mg olanzapine orally as 1 tablet or 1 tablet of 1 mg
risperidone followed by 180 ml water.
Blood samples (10 ml) were withdrawn before dos-
ing (time zero) and at time points of 0.5, 1.0, 1.5, 2.0,
3.0, 4.0, 5.0, 6.0, 8.0, 12.0, 24.0, and 48.0 hours after
dosing. Ten-milliliter syringes were used for this pur-
pose, and catheter lines were cleared of residual heparin
solution before sampling. The blood was transferred to
heparinized glass blood collection tubes (Vacutainer;
030-38markowitz119815.qxd 1/18/02 2:51 PM Page 32
32 Markowitz et al
Becton Dickinson, Rutherford, NJ) and centrifuged at
4°C for 5 minutes. The plasma was immediately aspi-
rated and transferred into polypropylene vials and
stored at –80°C until analysis. Cumulative urine was
collected for 24 hours after drug dosing, and the total
volume was recorded. A 25-ml aliquot of urine was
frozen at –80°C for subsequent analysis of olanzapine
and risperidone and their respective metabolites.
Quantitation of risperidone in plasma. All plasma con-
centration assays for risperidone were performed by
Pharma Bio-Research Group BV (GP Zuidlaren, The
Netherlands) with a validated liquid chromatography–
tandem mass spectroscopy method. After the addition
of 5 ng deuterated risperidone as an internal standard,
0.5-ml plasma samples were subjected to solid-phase
extraction. Then 3 ml of 0.1-mol/L phosphate buffer
(pH 6.0) was added to each sample. Vortex mixing was
performed, and each sample was poured over a previ-
ously conditioned solid phase extraction column. The
column was washed with 3 ml Milli-Q (Waters Inc,
Milford, Mass) water and 1 ml 1-mol/L acetic acid with
3 ml methanol. The column was eluted with 3 ml
methanol–25% ammonium hydroxide (98:2 [vol/vol]).
The eluates were then evaporated to dryness under
nitrogen.
A reversed-phase step gradient elution method was
used to separate the analytes on a C18 base deactivated
silica column (Hypersil BDS; Phenomenex, Torrance,
Calif) (10 cm × 4.6 mm inner diameter, packed with 3-
µm particles) with a 0.01-mol/L ammonium formate
(pH 4.0)/acetonitrile (67:33) eluent, followed by a step
gradient of 0.01-mol/L ammonium formate (pH
4.0)/acetonitrile (15:85) to wash the column. For detec-
tion, a triple quadrupole mass spectrometer (API 3000;
MDS Sciex, Concord, Ontario, Canada) was used in the
positive-ion mode. The lower limit of quantitation was
1 ng/ml risperidone. Interday and intraday coefficients
of variation (CVs) were <15%.
Quantitation of olanzapine in plasma. All olanzap-
ine plasma concentration assays were performed by
National Medical Services (Willow Grove, Pa) with a
validated gas chromatography method with nitrogen
and phosphorus detection. After the addition of the
internal standard (LY170222, a 2-ethylated olanzapine
derivative), plasma samples were subjected to auto-
mated solid phase extraction to isolate the analytes of
interest. A DB-17 analytical column was used to sepa-
rate the compounds of interest with a temperature ramp.
The lower limit of quantitation was 1 ng/ml olanzap-
ine. Interday and intraday CVs were <15%.
Quantitation of parent and conjugated metabolites
in urine. Urine samples were analyzed for risperidone
CLINICAL PHARMACOLOGY & THERAPEUTICS
JANUARY 2002
by means of the method previously described for
plasma risperidone. Aliquots (1 ml) of cumulative 24-
hour urine collections were analyzed for risperidone.
The concentration was multiplied by the total 0–24 hour
urinary volume to determine the amount of unchanged
risperidone excreted over this period. A second aliquot
of the 0–24 hour cumulative urine collection for each
subject was incubated with β-glucuronidase (2500 units
added to 1 ml 0.2-mol/L acetate buffer, pH 5) overnight
(16 hours) at 37°C to liberate β-glucuronidase–
sensitive conjugates, and the sample was analyzed for
risperidone. A third aliquot of each subject’s cumula-
tive 24-hour urine collection was subjected to acid
hydrolysis (1 ml 5N hydrochloric acid at 50°C) for 1
hour, neutralized (1 ml 5N sodium hydroxide), and ana-
lyzed for risperidone to determine the total amount of
conjugated and unconjugated risperidone in the urine
sample.
Urine samples were analyzed for olanzapine in the
investigator’s laboratory with a reversed phase HPLC
assay with ultraviolet detection.17 A 1-step liquid-liquid
extraction procedure was used to recover olanzapine and
the internal standard (LY170222) from urine. The 24-hour
urine sampling was conducted as described for risperi-
done and untreated urine, β-glucuronidase–incubated
urine, and acid-incubated urine assayed for olanzapine.
The lower limit of quantitation of the assay was 1 ng/ml
olanzapine. Interday and intraday CVs for olanzapine
were <15%. Recoveries of olanzapine and the internal
standard in urine were 79% ± 7% and 89% ± 7%,
respectively. Recoveries of conjugated drug were
unavailable because of the lack of authentic standards.
Therefore differences in extraction efficiencies caused
by the addition of buffer and acid in different incuba-
tion mixtures cannot be excluded. Accordingly, the
amounts of olanzapine excreted in the urine were only
compared within each urine preparation to obtain per-
centage differences between excretion of olanzapine
species with and without coadministration of
probenecid.
Data analysis
The pharmacokinetic data analysis program
P-Pharm (MicroPharm International, Champs, France)
was used to analyze plasma concentration–versus–
time data for olanzapine and risperidone. Both 1- and
2-term polyexponential equations with the addition of
absorption constant (ka) and lag time (tlag) functions as
estimated parameters were fitted to the data sets with
various combinations of concentration value weighting
and parameter distribution functions. Data were fitted
individually by means of an expectation–maximum
030-38markowitz119815.qxd 1/18/02 2:51 PM Page 33
CLINICAL PHARMACOLOGY & THERAPEUTICS
VOLUME 71, NUMBER 1
Fig 1. Plasma concentration-versus-time profile for risperi-
done administered alone (boxes) and coadministered with
probenecid (circles). Data are mean values ± SD (n = 12).
likelihood algorithm. The equation and weighting of
best fit were assessed according to the Akaike infor-
mation criterion, maximum likelihood, and visual
examination of residual values for systematic errors.
The area under the concentration-versus-time profile
(AUC) for various time periods was calculated with
standard methods.18 The maximum plasma concentra-
tion (Cmax) and the time to Cmax (tmax) were observed
directly from the data. Differences in the mean values
for the estimated pharmacokinetic parameters were
assessed with the paired 2-sided Student t test (SPSS
for Windows, Rel. 10.0.0; SPSS Science, Chicago,
Ill). The level of significance was set at P <.05.
RESULTS
Study subjects
Fourteen subjects (11 men, 3 women) were enrolled
in the study. Two subjects (1 man, 1 woman) discon-
tinued participation because of adverse experiences.
One subject withdrew from the study after initial
risperidone treatment because of dysphoria. The inves-
tigators asked the second subject to discontinue partic-
ipation after hypotension (113/30 mm Hg) and brady-
cardia (30 beats/min) developed after the initial dose
of olanzapine. She was placed in a recumbent position,
was given fluids orally, and recovered fully with vital
Markowitz et al 33
Fig 2. Plasma concentration-versus-time profile for olanza-
pine administered alone (boxes) and coadministered with
probenecid (circles). Data are mean values ± SD (n = 12).
Asterisk, Significantly different concentrations between
administration groups (P <.05).
signs returning to baseline values within 1 hour. Twelve
subjects completed all 4 treatment periods.
Risperidone pharmacokinetics
The plasma concentration-versus-time profiles for
risperidone are shown in Fig 1. A 1-term polyexponen-
tial equation with an absorption term and a lag phase
with homoscedastic weighting of plasma concentra-
tions was the equation that best fit the risperidone
plasma concentration-versus-time data. The primary
estimated pharmacokinetic parameters were apparent
oral clearance (CL/F), apparent oral volume of distrib-
ution (Vd/F), ka, and tlag. Individual parameters are
shown in Table I. Most pharmacokinetic parameters for
risperidone showed substantial variability between indi-
viduals and between the two administration phases.
However, there were no statistically significant differ-
ences in pharmacokinetic parameters for risperidone
between the two phases, and the mean plasma concen-
tration-versus-time profiles could almost be superim-
posed (Table I; Fig 1).
Olanzapine pharmacokinetics
Olanzapine plasma concentration-versus-time pro-
files are shown in Fig 2. A 2-term polyexponential
equation with an absorption term and a lag phase with
030-38markowitz119815.qxd 1/18/02 2:51 PM Page 34

CLINICAL PHARMACOLOGY & THERAPEUTICS

34 Markowitz et al JANUARY 2002

Table I. Pharmacokinetic parameters for risperidone after administration of risperidone alone and with probenecid
AUC0-24
CL/F (L/h) Vd/F (L) ka (h–1) tlag (h) (ng · ml/h) tmax (h) Cmax (ng/ml)

Subject R R+P R R+P R R+P R R+P R R+P R R+P R R+P

1 118 124 259 223 6.0 6.0 1.4 0.9 10.1 9.2 2.0 1.5 3.1 3.5
2 917 173 659 133 1.3 5.8 0.0 0.0 0.7 3.9 1.5 1.0 0.4 2.7
3 130 154 246 246 2.7 2.7 0.4 0.4 7.4 7.8 1.2 1.0 4.1 3.0
4 28 30 147 119 3.0 6.0 0.5 0.5 42.3 42.2 1.5 1.0 5.8 8.7
5 13 15 105 76 4.2 1.0 1.4 1.5 86.7 73.7 2.0 3.0 8.5 9.8
6 30 29 106 94 1.3 0.9 1.3 1.3 38.3 36.3 3.1 3.0 6.6 6.7
7 10 15 82 105 4.8 6.0 0.9 1.0 123.1 91.0 1.7 1.5 12.1 9.4
8 56 62 130 197 6.0 3.0 0.3 0.6 20.4 17.5 1.0 1.5 6.2 3.9
9 44 113 133 325 1.7 1.4 1.9 2.4 24.1 11.4 3.0 4.0 4.9 1.9
10 72 82 141 211 4.0 6.0 1.3 1.3 16.2 13.7 1.5 2.0 4.2 3.8
11 16 8 72 86 2.0 1.3 1.3 1.2 64.6 129.7 3.0 4.2 9.9 9.8
12 33 39 101 107 2.6 2.4 1.4 1.4 32.3 28.7 1.9 2.0 6.6 6.2
Mean value 122 70 182 160 3.3 3.5 1.0 1.0 38.8 38.8 1.9 2.1 6.0 5.8
SD 253 58 161 79 1.7 2.2 0.6 0.6 36.2 39.6 0.7 1.1 3.1 3.0
P value* NS NS NS NS NS NS NS

R, Risperidone alone; R + P, risperidone plus probenicid; CL/F, apparent oral clearance; Vd/F, apparent oral volume of distribution; ka, absorption constant; tlag, lag
time; AUC0-24, area under the plasma concentration-versus-time curve; tmax, time to reach Cmax; Cmax, maximum plasma concentration; NS, no significant difference
between risperidone alone and risperidone with probenecid (P > .05).
*Paired 2-sided Student t test.

homoscedastic weighting of plasma concentrations was
the equation that best fit the olanzapine plasma concen-
tration-versus-time data. The primary estimated phar-
macokinetic parameters were CL/F and Vd/F. These
values are shown in Table II. Most pharmacokinetic
parameters for olanzapine showed substantial variabil-
ity between individuals and between the two adminis-
tration phases. Statistically significant differences
between pharmacokinetic parameters for olanzapine
administered alone and coadministered with probenecid
were observed with the probenecid phase showing a
faster absorption rate and higher Cmax (Table II; Fig 2).
Urinary metabolites
There were no significant differences in the amount
of risperidone detected in any of the 24-hour urine spec-
imens. Neither β-glucuronidase treatment nor acid
incubation led to any appreciable changes in the amount
of risperidone recovered from urine.
The urinary olanzapine data showed no significant
difference in the amount of unchanged olanzapine
excreted over the first 24 hours (P >.05; Table III). Sim-
ilarly, the acid-treated urine samples, indicative of the
total amount of olanzapine conjugation, also showed
no significant differences between the administration
phases (P > .05). However, the β-glucuronidase–treated
samples showed a statistically significant difference
between phases, the amount of β-glucuronidase–sensi-
tive olanzapine metabolite being significantly reduced
(P < .001) in the probenecid phase in most cases. A sig-
nificant difference was observed between untreated and
β-glucuronidase–treated samples after administration
of olanzapine alone (P = .0013), but no significant dif-
ference was observed between these groups of samples
after coadministration of probenecid.
DISCUSSION
The most important finding was that probenecid had
no effect on the pharmacokinetics of risperidone but
had several significant effects on several pharmacoki-
netic measures for olanzapine. The observed statisti-
cally significant increases in olanzapine AUC0-24h and
Cmax and the increase in ka after coadministration of
probenecid (Table II) were consistent with an overall
reduced presystemic metabolism of olanzapine,
although the effect was not consistent for each individ-
ual. This effect was presumably caused by a reduced
capacity for glucuronidation of olanzapine when
probenecid was coadministered. The lack of effect on
CL/F suggests inhibition of hepatic glucuronidation by
probenecid was not of sufficient magnitude to change
the overall clearance of a single dose of olanzapine. It
should be emphasized that these data are from a single-
dose study design and not under steady-state condi-
tions, so the clinical relevance of these findings is
unknown.
The data on urinary excretion of olanzapine showed
no average difference in the amount of unchanged olan-
030-38markowitz119815.qxd 1/18/02 2:52 PM Page 35
CLINICAL PHARMACOLOGY & THERAPEUTICS
VOLUME 71, NUMBER 1
Fig 3. Glucuronidation pathways of olanzapine. UDPGA, Uridine diphosphate glucuronic acid; UDP, uridine
diphosphoglucuronate; UDPGT, uridine diphosphoglucuronate–glucuronosyltransferase; OLZ, olanzapine.
zapine recovered in the urine over the first 24 postdose
hours between olanzapine alone and olanzapine coad-
ministered with probenecid. β-Glucuronidase has been
reported to liberate free olanzapine from the 4´-N-
glucuronide conjugate, but the non–β-glucuronidase–
sensitive conjugate appears resistant to cleavage by this
enzyme (Fig 3).19 When the cumulative 24-hour urine
samples were incubated with β-glucuronidase, there
was only 46% excretion of olanzapine compared with
the condition of olanzapine administered alone (Table
III). This finding suggested that the 4´-N-glucuronide
conjugate accounts for a substantial amount of the olan-
zapine species excreted into urine. In contrast, coad-
ministration of olanzapine with probenecid showed
only a modest and insignificant mean increase in olan-
zapine concentration (115%; Table III) after acid treat-
ment. This finding suggested activity of the pathway
that produces the 4´-N-glucuronide conjugate was sub-
stantially inhibited by probenecid.
When the 24-hour cumulative urine samples were
treated with acid to nonspecifically cleave all conju-
gates, there was no difference in the olanzapine recov-
ered. These data suggest that the total amounts of olan-
zapine absorbed and excreted, either as unchanged
Markowitz et al 35
olanzapine or as conjugated metabolites, were the same
and that probenecid may not substantially affect forma-
tion of the 10-N-olanzapine conjugate glucuronidation
pathway. Although olanzapine glucuronides were not
directly measured, the results from the urinary excre-
tion analysis suggest that probenecid inhibits formation
of 4´-N-olanzapine glucuronide and that this inhibition
may account for the differences in olanzapine plasma
pharmacokinetics observed between the two drug
administration phases. Some caution is appropriate in
interpreting these data because identical recoveries of
olanzapine from β-glucuronidase and acid-treated urine
cannot be assured.
Compounds structurally similar to olanzapine, such
as clozapine and loxapine, are known to form quater-
nary ammonium–linked glucuronides mediated by
UGT1A4.20,21 These results suggest that probenecid
may inhibit the UGT1A4 pathway and that the inhibi-
tion may lead to formation of the 4´-N-olanzapine glu-
curonide metabolite while sparing the conjugation of
species such as 10-N-glucuronide, which may be
formed through alternate enzyme pathways (Fig 3). An
additional consideration in interpretation of our results
is the potential effect of probenecid on P-glycoprotein
030-38markowitz119815.qxd 1/18/02 2:52 PM Page 36

CLINICAL PHARMACOLOGY & THERAPEUTICS

36 Markowitz et al JANUARY 2002

Table II. Pharmacokinetic parameters for olanzapine after administration of olanzapine alone and with probenecid
AUC0-24 Cmax
CL / F (L/h) Vd / F (L) ka (h–1) tlag (h) (ng · ml/ h) tmax (h) (ng/ml)

Subject O O+P O O+P O O+P O O+P O O+P O O+P O O+P

1 31.2 54.3 285 725 1.07 2.38 2.7 2.0 16 76 4.1 3.0 5.4 6.0
2 151.1 60.2 755 911 2.47 0.65 3.8 2.0 32 58 6.0 6.0 4.2 4.4
3 44.1 27.6 274 460 0.80 0.66 0.4 1.8 107 128 1.8 8.0 12 9.3
4 18.9 13.0 210 559 0.35 2.75 3.0 0.8 109 147 8.0 3.0 6.9 8.4
5 11.6 9.5 557 661 0.22 1.69 1.3 1.5 99 144 8.0 4.0 5.8 10
6 2.0 21.8 555 519 0.61 0.87 1.2 2.7 52 105 4.0 6.0 5.3 7.8
7 22.4 28.3 788 100 0.14 0.670 0.0 0.9 83 83 8.0 4.0 4.9 5.0
8 18.1 21.8 744 531 1.35 2.79 0.8 1.9 115 124 8.1 5.0 6.7 9.3
9 5.5 8.8 589 388 0.40 0.53 1.5 0.7 106 149 5.0 5.0 6.2 8.8
10 11.1 11.8 461 227 1.11 0.91 1.4 1.2 132 144 3.0 3.0 7.3 11.0
11 6.8 11.3 464 365 4.00 5.00 1.6 2.0 195 183 2.3 3.0 11.0 11.0
12 2.1 21.9 405 534 0.12 0.86 0.6 1.9 95 107 11.7 6.0 5.7 6.7
Mean value 27.1 24.2 507 498 1.05 1.65 1.5 1.6 95 120 5.8 4.7 6.8 8.1
SD 41.0 16.9 194 217 1.14 1.36 1.1 0.6 47 36 3.0 1.6 2.4 2.2
P value* NS NS .024 NS .002 NS .024

O, Olanzapine alone; O + P, olanzapine plus probenecid; NS, no significant difference between olanzapine alone and olanzapine with probenecid (P > .05). Other
abbreviations as in Table I.
*Paired 2-sided Student t test.

Table III. Urinary excretion (0–24 hours) of unchanged olanzapine (untreated), β-glucuronidase–sensitive metabo-
lites, and acid-sensitive metabolites in healthy volunteers coadminsitered olanzapine and probenecid relative to
amounts of olanzapine excreted when olanzapine was administered alone
Subject No. Untreated urine β-Glucuronidase–treated urine Acid-treated urine

1 50 135 100
2 87 41 121
3 103 52 68
4 213 51 69
5 135 13 240
6 121 118 107
7 68 22 100
8 112 23 119
9 151 57 126
10 93 15 129
11 120 9 103
12 115 12 93
Mean value 114 46 115
SD 42 41 44
P value* NS .001 NS
Data are percentage excretion compared with olanzapine administered alone. NS, No significant difference between olanzapine alone and olanzapine with probenecid
(P >.05).
*Paired 2-sided Student t test.

because this drug is known be a potent inhibitor of this
protein activity.22 However, results of preliminary in
vitro studies suggest that olanzapine may be a poor sub-
strate for P-glycoprotein and that risperidone appears
to be a good substrate.23
Glucuronidation has been extensively studied in
vitro, and at least 33 families of UDP-UDPGT have
been identified.11 Interindividual differences of up to
10-fold in UDPGT activity are known to exist. Once
formed, glucuronide conjugates can be enzymatically
hydrolyzed to liberate the parent drug, which can be
pharmacologically active and which can then be recon-
jugated or excreted. Factors including smoking behav-
ior, sex, obesity, genetic polymorphism, and concomi-
030-38markowitz119815.qxd 1/18/02 2:52 PM Page 37
CLINICAL PHARMACOLOGY & THERAPEUTICS
VOLUME 71, NUMBER 1
tant medications are known to affect glucuronidation
and hence eliminate various medications that are pri-
marily UDPGT substrates.24,25
The role of inhibition of phase 2 metabolism as a
mechanism of drug interactions involving psychotropic
drugs has received little study. In this study probenecid
was used as an inhibitor of glucuronidation. The dosage
used, 500 mg probenecid twice daily, is typical in the
management of gout. However, doses may be titrated
up to 2 g/d for some conditions. Thus it is possible that
the observed changes in olanzapine plasma disposition
would be of greater magnitude depending on the con-
dition for which a patient is treated (dose dependent).
Although this study was designed to assess the general
effects of probenecid as an inhibitor compound, numer-
ous other psychotropic drugs undergo extensive glu-
curonidation. Some, such as oxazepam26 and valproic
acid,27 have been implicated as inhibitors of glu-
curonidation and are commonly coadministered with
antipsychotics such as risperidone and olanzapine.
Numerous other medications, such as nonsteroidal anti-
inflammatory drugs, inhibit glucuronidation in vitro
and hence may alter the disposition of other coadmin-
istered medications. Genetic polymorphism has been
found for a number of UGTs, including UGT1A4, an
enzyme that may contribute to the metabolism of olan-
zapine.16 The variability in dosing of atypical antipsy-
chotic drugs is an important factor affecting the selec-
tion of specific medications for clinical use, cost, and
overall utilization. Although the clinical significance of
phase 2 metabolic inhibition of olanzapine cannot be
predicted from these data, identification of coadminis-
tered drugs that interact with antipsychotics during rou-
tine clinical care remains a priority in drug interaction
studies in psychiatry.
References
1. Markowitz JS, Brown CS, Moore TR. Atypical antipsy-
chotics. I. Pharmacology, pharmacokinetics, and efficacy.
Ann Pharmacother 1999;33:73-85.
2. Preskorn SH, Burke MJ, Fast GA. Therapeutic drug mon-
itoring: principles and practice. Psychiatr Clin North Am
1993;16:611-41.
3. Cooper TB. Clozapine plasma level monitoring: current
status. Psychiatr Q 1996;67:297-311.
4. Perry PJ, Sanger T, Beasley C. Olanzapine plasma con-
centrations and clinical response in acutely ill schizo-
phrenic patients. J Clin Psychopharmacol 1997;17:
472-7.
5. DeVane CL, Markowitz JS. Drugs as substrates of meta-
bolic enzymes: antipsychotics. In: Levy RH, Thummel
KE, Trager WF, editors. Metabolic drug interactions.
Philadelphia: Lippincott Williams & Wilkins; 2000.
Markowitz et al 37
6. Llerena A, Dahl ML, Ekqvist B, Bertilsson L. Haloperi-
dol disposition is dependent on the debrisoquine hydrox-
ylation phenotype: increased plasma levels of the reduced
metabolite in poor metabolizers. Ther Drug Monit
1992;14:261-4.
7. Lane HY, Hu OY, Jann MW, Deng HC, Lin HN, Chang
WH. Dextromethorphan phenotyping and haloperidol dis-
position in schizophrenic patients. Psychiatry Res
1997;69:105-11.
8. Bertilsson L, Carrillo JA, Dahl ML, Llerena A, Alm C,
Bondesson U, et al. Clozapine disposition covaries with
CYP1A2 activity determined by the caffeine test. Br J
Clin Pharmacol 1994;38:471-3.
9. Dahl ML, Llerena A, Bondesson U, Lindström L, Bertils-
son L. Disposition of clozapine in man: lack of an asso-
ciation with debrisoquine and S-mephenytoin hydroxyla-
tion polymorphisms. Br J Clin Pharmacol 1994;37:71-4.
10. Arranz MJ, Dawson E, Shaikh S, Sham P, Sharma T,
Aitchison K, et al. Cytochrome P4502D6 does not deter-
mine response to clozapine. Br J Clin Pharmacol
1995;39:417-20.
11. Mackenzie PI, Owens IS, Burchell B, Bock KW, Bairoch A,
Belanger A, et al. The UDP glycosyltransferase gene
superfamily: recommended nomenclature update based
on evolutionary divergence. Pharmacogenetics 1997;
7:255-69.
12. Kassahun K, Mattiuz E, Nyhart E Jr, Obermeyer B,
Gillespie T, Murphy A, et al. Disposition and biotrans-
formation of the antipsychotic agent olanzapine in
humans. Drug Metab Dispos 1997;25:81-93.
13. DeVane CL, Nemeroff CB. An evaluation of risperidone
drug interactions. J Clin Psychopharmacol 2001;21:408-16.
14. Callaghan JT, Berstrom RF, Ptak LR, Beasley CM. Olan-
zapine: pharmacokinetic and pharmacodynamic profile.
Clin Pharmacokinet 1999;37:177-93.
15. Abernethy DR, Greenblatt DJ, Ameer B, Shader RI.
Probenecid impairment of acetaminophen and lorazepam
clearance: direct inhibition of ether glucuronide forma-
tion. J Pharmacol Exp Ther 1985;234:345-9.
16. Liston H, Markowitz JS, DeVane CL. Drug glucuronida-
tion in clinical psychopharmacology. J Clin Psychophar-
macol 2001;21:500-15.
17. Boulton DW, Markowitz JS, DeVane CL. A high per-
formance liquid chromatography assay with ultravio-
let detection for olanzapine in human plasma and
urine. J Chromatogr B Biomed Sci Appl 2001;759:
319-23.
18. Rowland M, Tozer TN. Clinical pharmacokinetics: con-
cepts and applications. 2nd ed. London: Lea and Febiger;
1989.
19. Kassahun K, Mattiuz E, Franklin R, Gillespie T. Olanza-
pine 10-N-glucuronide-A tertiary N-glucuronide unique
to humans. Drug Metab Dispos 1998;26:848-55.
20. Green MD, Tephly TR. Glucuronidation of amines and
hydroxylated xenobiotics and endobiotics catalyzed by
expressed human UGT1.4 protein. Drug Metab Dispos
1996;24:356-63.
030-38markowitz119815.qxd 1/18/02 2:52 PM Page 38

CLINICAL PHARMACOLOGY & THERAPEUTICS

38 Markowitz et al JANUARY 2002

21. Green MD, Bishop WP, Tephly TR. Expressed human
UGT1.4 protein catalyzes the formation of quaternary
ammonium-linked glucuronides. Drug Metab Dispos
1995;23:299-302.
22. Ward ES, Pollack GM, Brouwer KLR. Probenecid-asso-
ciated alterations in valproic acid pharmacokinetics in
rats: can in vivo disposition of valproate glucuronide be
predicted from in vitro formation data? Drug Metab Dis-
pos 2000;28:1433-9.
23. DeVane CL, Boulton DW, Liston HL, Markowitz JS.
Human recombinant P-glycoprotein specificity for
antipsychotics. In: Proceedings of the Thirty-ninth
Annual Meeting of the American College of Neuropsy-
chopharmacology; 2000 Dec 9-12; San Juan, Puerto Rico.
San Juan: The College; 2000.
24. Miners JO, Mackenzie PI. Drug glucuronidation in
humans. Pharmacol Ther 1991:51:347-69.
25. Kroemer HK, Klotz U. Glucuronidation of drugs: a re-
evaluation of the pharmacological significance of the
conjugates and modulating factors. Clin Pharmacokinet
1992:23:292-310.
26. Wahlström A, Pacifici GM, Lindström B, Hammar L,
Rane A. Human liver morphine UDP-glucuronyltrans-
ferase enantioselectivity and inhibition by opioid con-
geners and oxazepam. Br J Pharmacol 1988;94:
864-70.
27. Samara EE, Granneman RG, Witt GF, Cavanaugh JH.
Effect of valproate on the pharmacokinetics and pharma-
codynamics of lorazepam. J Clin Pharmacol 1997;
37:442-50.

Receive tables of contents by e-mail

To receive the tables of contents by e-mail, sign up through our Web site at
http://www.mosby.com/cpt
Choose “E-mail Notification.”
Simply type your e-mail address in the box and click the “Subscribe” button.

Alternatively, you may send an e-mail message to majordomo@mosby.com. Leave the subject line
blank and type the following as the body of your message:
subscribe cpt_toc
You will receive an e-mail to confirm that you have been added to the mailing list. Note that
table of contents e-mails will be sent out when a new issue is posted to the Web site.