Simultaneous determination of etoposide and its catechol metabolite in the plasma of pediatric patients by liquid chromatography/tandem mass spectrometry

Article abstract of DOI:10.1002/jms.173

The anticancer drug etoposide is associated with leukemias with MLL gene translocations and other translocations as a treatment complication. The genotype of cytochrome P450 3A4 (CYP3A4), which converts etoposide to its catechol metabolite, influences the

Full text of DOI:10.1002/jms.173

JOURNAL OF MASS SPECTROMETRY
J. Mass Spectrom. 2001; 36: 771–781
Simultaneous determination of etoposide and its
catechol metabolite in the plasma of pediatric patients
by liquid chromatography/tandem mass spectrometry
Shaokun Pang,1† Naiyu Zheng,1 Carolyn A. Felix,2 Jennifer Scavuzzo,2 Ray Boston3 and Ian
A. Blair1∗
1
Center for Cancer Pharmacology, Department of Pharmacology, University of Pennsylvania, Philadelphia, Pennsylvania 19104-6160, USA
Division of Oncology, The Children’s Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-
2
4318, USA
3
Biomathematics Unit, School of Veterinary Medicine, University of Pennsylvania, Kennett Square, Pennsylvania 19348, USA
Received 4 December 2000; Accepted 16 March 2001
The anticancer drug etoposide is associated with leukemias with MLL gene translocations and other
translocations as a treatment complication. The genotype of cytochrome P450 3A4 (CYP3A4), which
converts etoposide to its catechol metabolite, influences the risk. In order to perform pharmacokinetic
studies aimed at further elucidation of the translocation mechanism, we have developed and validated
a liquid chromatography/electrospray/tandem mass spectrometry assay for the simultaneous analysis of
etoposide and its catechol metabolite in human plasma. The etoposide analog teniposide was used as
the internal standard. Liquid chromatography was performed on a YMC ODS-AQ column. Simultaneous
determination of etoposide and its catechol metabolite was achieved using a small volume of plasma,
so that the method is suitable for pediatric patients. The limits of detection were 200 ng ml−1 etoposide
and 10 ng ml−1 catechol metabolite in human plasma and 25 ng ml−1 etoposide and 2.5 ng ml−1 catechol
metabolite in protein-free plasma, respectively. Acceptable precision and accuracy were obtained for
concentrations in the calibration curve ranges 0.2–100 µg ml−1 etoposide and 10–5000 ng ml−1 catechol
metabolite in human plasma. Acceptable precision and accuracy for protein-free human plasma in the range
25–15 000 ng ml−1 etoposide and 2.5–1500 ng ml−1 etoposide catechol were also achieved. This method
was selective and sensitive enough for the simultaneous quantitation of etoposide and its catechol as a
total and protein-free fraction in small plasma volumes from pediatric cancer patients receiving etoposide
chemotherapy. A pharmacokinetic model has been developed for future studies in large populations.
Copyright  2001 John Wiley & Sons, Ltd.
KEYWORDS: etoposide; catechol metabolite; plasma; high-performance liquid chromatography; tandem mass spectrometry
(Scheme 1).2 Etoposide catechol can undergo sequential one-
electron oxidations to form a semiquinone and a quinone
(Scheme 1).3,4 Subsequent redox cycling results in the gen-
eration of reactive oxygen species (ROS), such as hydrogen
peroxide and superoxide,4 which undergo Fenton chem-
istry to form hydroxyl radicals. The hydroxyl radicals
either damage DNA directly5 or induce the formation of
lipid hydroperoxides, which break down to DNA-reactive
aldehydic bifunctional electrophiles, malondialdehyde, 4-
hydroxy-2-nonenal, and 4-oxo-2-nonenal (Scheme 1).6 In
addition, etoposide semiquinone and quinone can directly
form adducts with DNA and protein molecules.4,7–10 The
primary mechanism of cytotoxicity of etoposide involves
DNA topoisomerase II-mediated chromosomal breakage.11,12
DNA topoisomerase II mediates essential changes in DNA
topology by transient cleavage and religation of the double
helix. Etoposide stabilizes the DNA cleavage and covalently
linked enzyme intermediate by decreasing religation.11,12
The resultant chromosomal breakage initiates apoptosis.13
INTRODUCTION
Etoposide (Fig. 1) is a semi-synthetic glycoside derivative
of podophyllotoxin, originally an extract of the man-
drake plant. It is used for the treatment of a variety
of malignancies, such as small-cell lung cancer, testicu-
lar carcinoma, lymphoma, other solid tumors and sev-
eral types of leukemia.1 Teniposide, a structurally related
epipodophyllotoxin analog of etoposide, has also been used
as an anticancer drug (Fig. 1). One of the major path-
ways of etoposide metabolism involves cytochrome P450
3A4 (CYP3A4)-mediated formation of a catechol metabolite
ŁCorrespondence to: I. A. Blair, Center for Cancer Pharmacology,
Department of Pharmacology, University of Pennsylvania,
Philadelphia, Pennsylvania 19104-6160, USA.
E-mail: ian@spirit.gcrc.upenn.edu
†Present address: Purdue Pharma, L.P., 444 Saw Mill River Road,
Ardsley, New York 10502, USA.
Contract/grant sponsor: NIH; Contract/grant number: CA77683;
MO1-RR00240; UO1 5-U01-HD-37255.
DOI: 10.1002/jms.173
Copyright  2001 John Wiley & Sons, Ltd.

772
S. Pang et al.
ROS and quinone formation, which results from etoposide
metabolism has been suggested as an alternative mechanism
for DNA damage.4,7 The same mechanisms may be relevant
to leukemia-associated chromosomal translocations.
MLL gene at chromosome band 11q23, as a treatment
complication.15 The genotype of CYP3A4, which converts
etoposide to its O-demethylated catechol metabolite, influ-
ences the risk.15 Recently, we showed that etoposide catechol
and etoposide quinone damage the MLL gene in a DNA-
topoisomerase II-dependent, sequence-specific manner.16
Furthermore, the N-7 position of guanine was critical to this
damage, suggesting that it was a result of alkylation by the
metabolites. It is not known whether individual differences
in pharmacokinetics of etoposide or of its catechol metabo-
lite, or both, are relevant to the leukemogenesis. Therefore,
it is important to understand the contribution of the parent
drug and its metabolites to the chromosomal breakage.
The objective of the present study was to develop a phar-
macokinetic model that can ultimately be used to understand
better the role of etoposide and its metabolites in the translo-
cation process. Inter-patient differences in etoposide and
etoposide catechol pharmacokinetics may provide insights
into the CYP3A4 genotype–leukemia association. Pharma-
cokinetic monitoring may permit rational optimization of
etoposide dose and schedule so that the risk of leukemia
is reduced. Methods used previously for the determination
of etoposide in human plasma include, high-performance
liquid chromatography (HPLC) with UV detection,17 flu-
orescence detection18,19 and electrochemical detection.20–23
HPLC methods based on electrochemical detection also
have been used for the analysis of etoposide catechol in
Etoposide is a substrate for metabolism by CYP3A4.14
Therapy with etoposide is associated with leukemias with
chromosomal translocations, especially translocations of the
H
O
O
HO
R2
O
O
OH
O
O
O
O
H3CO
OR1
OH
Etoposide (MW=588)
Etoposide catechol (MW=574) R1= H
Teniposide (MW=656) R1= CH3
R1= CH3
R2= CH3
R2= CH3
R2= C4H3S
Figure 1. Structures of etoposide, etoposide catechol and
teniposide (internal standard).
H
H
O
O
O
HO
O
HO
H3C
O
H3C
O
OH
O
O
OH
O
O
CYP3A
O
O
O
O
O
O
H3CO
OH
H3CO
etoposide
OCH3
OH
OH
.
--
catechol
-
O2
.
O2
H
HO
O
O
HO
H3C
O
O
OH
O
2e/ 2H+
O
oxidative
DNA damage
lipid
hydroperoxides
O
O
H3CO
O
O
-
O2
O2
.
.
-
malondialdehyde
4-hydroxy-2-nonenal
4-oxo-2-nonenal
HO
H
O
O
HO
O
H3C
O
OH
O
O
O
O
DNA-adducts
H3CO
quinone
O
O
Scheme 1. Formation of etoposide catechol, etoposide semiquinone and etoposide quinone by metabolism of etoposide.
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781

Determination of etoposide and its catechol in plasma
773
human plasma.23,24 In patients with cancer, a substantial
amount of plasma etoposide is non-covalently bound to
plasma proteins with significant amount of inter-patient
variation.25 Protein binding, which varied from 80% to
97% (mean, 93%) for different individuals, appeared to be
dependent upon serum albumin concentrations.25 However,
it is still unknown whether there are inter-individual dif-
ferences in etoposide catechol binding. Furthermore, it is
unclear whether there are changes in binding of etopo-
side and its catechol metabolite during multiple dosing
regimens. The unbound (protein-free) fraction of etoposide
correlates more closely with both toxicity and efficacy than
the total drug (protein-free C non-covalently protein bound)
concentration.26 The concentrations of protein-free etoposide
and protein-free etoposide catechol should be much lower
than the total plasma concentrations. Therefore, highly sen-
sitive analytical methodology with a wide dynamic range is
necessary if both protein-free and total drug concentrations
are monitored. We have developed a method with HPLC
combined with electrospray ionization tandem mass spec-
trometry (LC/ESI-MS/MS) that simultaneously quantifies
etoposide and its catechol metabolite as protein-free and total
concentrations in human plasma samples. The method was
successfully applied to pharmacokinetic studies in pediatric
cancer patients receiving multiple-day intravenous infusions
of etoposide.
for 10 min, the supernatant was discarded and the residue
was washed with diethyl ether three times. Pure etoposide
quinone was kept inside an aluminum foil-wrapped flask,
dried under vacuum overnight and stored at room tempera-
ture in the dark. It was converted to etoposide catechol by the
method of Relling et al.28 Briefly, etoposide quinone (0.296 g,
0.505 mmol) was dissolved in dioxane (5 ml) and water
(10 ml) under an atmosphere of nitrogen (sealed Aldrich
AtmosBag). A solution of 1 M sodium borohydride (NaBH4;
2 ml) was then added dropwise with stirring. The solution
was kept on an ice–water bath and more 1 M NaBH4 solu-
tion (0.8 ml) was added dropwise. After 10 min, the solution
was treated with 1.2 ml of 1 M HCl to stop the reaction. The
reaction mixture was extracted with dichloromethane (4 ml
once, and then 4 ð 2 ml), the organic phases were combined,
washed with water (5 ð 2 ml), and dried over Na2SO4. After
evaporation to dryness under a stream of nitrogen, the
residue was recrystallized with dichloromethane–diethyl
ether as described for etoposide quinone. Recrystallized
etoposide catechol was stored in the dark at room tem-
perature.
Liquid chromatography
HPLC separation was performed on a Waters (Milford,
MA, USA) model 2690 Alliance HPLC system with a YMC
(Wilmington, NC, USA) ODS-AQ analytical column (150 ð
2.0 mm i.d., 3 µm) preceded by a precolumn filter (2 µm;
Alltech, Deerfield, IL, USA). The mobile phases consisted
of 5 mM HCOONH4 and 0.1% aqueous formic acid solution
with 10% acetonitrile as A and 90% acetonitrile as B. The
analytes were eluted with a gradient at a flow-rate of 0.2 ml
minꢀ1. A set of 50 samples (200 µl of each) was run at one time
EXPERIMENTAL
Chemicals and materials
Etoposide was purchased from Sigma (St. Louis, MO, USA).
Teniposide was a gift from Bristol-Myers Squibb (Prince-
ton, NJ, USA). Ascorbic acid, ammonium formate, diox-
ane, sodium metaperiodate and sodium borohydride were
obtained from Aldrich (Milwaukee, WI, USA). HPLC-grade
water, formicacid, sodiumsulfateanddichloromethane were
obtained from Fisher Scientific (Fair Lawn, NJ, USA). HPLC-
grade acetonitrile was obtained from B&J (Muskegon, MI,
USA). Blank human plasma and protein-free plasma were
purchased from Biological Specialty (Landsdale, PA, USA).
Diethyl ether was obtained from VWR (Bridgeport, NJ, USA),
and ammonium sulfate from Mallinckrodt (Paris, KY, USA).
°
with the autosampler maintained at 5 C. When the samples
were analyzed, the column effluent was diverted to waste
for the first 2 min and the last 10 min in order to minimize
contamination of the mass spectrometer.
Mass spectrometry
ESI-MS analysis was performed using a Finnigan TSQ7000
triple-quadrupole mass spectrometer (ThermoQuest, San
Jose, CA, USA) equipped with a Finnigan electrospray
ionization source. The mass spectrometer was operated in
the positive ion mode and nitrogen was used as both sheath
gas (70 psi) and auxiliary gas (25 units). The source was
Synthesis of etoposide catechol
(epipodophyllotoxin catechol glucoside)
Etoposide quinone was synthesized according to the
patent of Nemec27 with minor modifications. Etopo-
side (0.130 g, 0.221 mmol) was dissolved in dioxane (2.4 ml)
and water (4.7 ml). One portion of 1.4 ml of 0.5 M sodium
metaperiodate (NaIO4) was added with stirring to this solu-
°
maintained at 200 C and the needle potential at 4.5 kV. The
multiplier voltage was set at 1500 V and argon was used as
the collision gas at 3.5 mTorr (1 Torr D 133.3 pa). The mass
spectrometer, configured in the selected reaction monitoring
(SRM) scan mode, monitored the transitions m/z 592.2 !
229.2 for etoposide catechol and m/z 606.2 ! 229.2 for
etoposide from 0.0 to 7.0 min, and m/z 674.2 ! 229.2 for
teniposide (internal standard) from 7.0 to 10.0 min with
a scan time of 0.5 s. The collision energy was optimized
to ꢀ32 eV for the determination of both etoposide and
teniposide and to ꢀ25 eV for the determination of etoposide
catechol. The data system was controlled using Finnigan
ICIS 8.3.0 software and Finnigan LCQuan V1.2 was used to
calculate peak areas.
°
tion. The reaction mixture was kept in the dark at 10 š 4 C
for 110 min then saturated with solid (NH4)2SO4. Etopo-
side quinone was extracted with dichloromethane (1 ð 4 ml,
then 5 ð 2 ml). The organic phases were combined, washed
with water (5 ð 2 ml), and dried with Na2SO4. After evap-
oration to dryness under a stream of nitrogen, the residue
was dissolved in a minimum volume of dichloromethane.
Diethyl ether was added to the dichloromethane solution
to initiate crystallization. After centrifugation at 3000 rpm
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781

774
S. Pang et al.
Catechol stability
etoposide and 10 ng mlꢀ1 –5 µg mlꢀ1 etoposide catechol for
total (bound C free) analysis. For the determination of
protein-free concentrations, known amounts of etoposide
and etoposide catechol were added to drug- and protein-free
human plasma at concentrations of 25 ng mlꢀ1 –15 µg mlꢀ1
for etoposide and at 2.5 ng mlꢀ1 –1500 ng mlꢀ1 for etoposide
catechol. The data were analyzed by using LCQuan Ver. 1.2.
Quantitation was achieved by plotting the peak area ratios
of the analytes to the internal standard versus concentration
followed by linear regression analysis with a weighting factor
of 1/x2. A water blank, plasma blank and control sample
(blank plasma spiked with internal standard only) were also
prepared and analyzed with each calibration curve.
In order to test the stability of the analytes in the injection
solutions, etoposide, etoposide catechol and teniposide were
spiked into 34% aqueous acetonitrile solutions containing 0,
°
0.2, 2 and 50 mM ascorbic acid. Samples were kept at 5 C and
taken at certain time points for analysis. In order to test the
stability of etoposide catechol during the sample preparation,
etoposide catechol was spiked into blank plasma and left at
room temperature. Samples were taken at 0, 0.5, 1, 2, 3 and
4 h for analysis.
Sample preparation
Etoposide, etoposide catechol and teniposide were dissolved
in methanol. Working standard solutions were prepared
by diluting the stock standard solution (1 mg mlꢀ1) with
methanol. Concentrations of the final plasma standard
solutions were one hundredth of the working standard
concentrations so that the final concentration of methanol
was only 1% of the total volume in plasma samples. Standard
plasma samples were prepared by spiking etoposide and
etoposide catechol into blank human plasma. All the
Accuracy and precision
The method for the analysis of both total and protein-free
fractions of etoposide and its catechol in human plasma
was validated by analyzing the quality control (QC) samples
on three separate days. For total (bound C free) etoposide
and etoposide catechol determination, the QC samples
were prepared by adding known amount of etoposide and
etoposide catechol to drug-free human plasma. The QC
samples contained a ratio of etoposide to etoposide catechol
of 20 : 1 with concentrations of 200 and 10 ng mlꢀ1 (lower
limit of quantitation, LLQ), 500 and 25 ng mlꢀ1 (lower quality
control, LQC), 5000 and 250 ng mlꢀ1 (middle quality control,
MQC) and 50 000 and 2500 ng mlꢀ1 (high quality control,
HQC). For protein-free etoposide and etoposide catechol,
the QC samples were prepared by adding known amount
of etoposide and etoposide catechol to drug- and protein-
free human plasma. The QC samples contained a ratio of
etoposide to etoposide catechol of 10 : 1 with concentrations
of 25 and 2.5 ng mlꢀ1 (LLQ), 200 and 10 ng mlꢀ1 (LQC),
1000 and 100 ng mlꢀ1 (MQC) and 10 000 and 1000 ng mlꢀ1
(HQC). The QC samples were analyzed on three separate
days and the LLQ samples were analyzed on one day. The
accuracy of this method was determined by comparing the
means of the measured concentrations with the theoretical
concentrations in the quality control samples and presented
as a percentage. The intra-day precision was expressed as the
relative standard deviation (RSD) of the sample replicates
over their mean values at each concentration within the same
validation day. The inter-day precision was expressed as the
RSD of three different validation days.
°
standard samples were aliquoted and stored at ꢀ80 C until
they were used. The concentration of the internal standard
(teniposide) was 10 µg mlꢀ1. All plasma (regular or protein-
free) standards and patient samples were thawed on ice
before analysis. For the quantitation of total etoposide and its
catechol in human plasma samples, plasma standard solution
(50 µl), teniposide (internal standard, 20 µl, 10 µg mlꢀ1),
freshly made ascorbic acid solution (20 µl, 50 mM) and
acetonitrile (260 µl) were added to an Eppendorf centrifuge
tube. After vortex mixing for 5 min, the sample tubes were
centrifuged in an Eppendorf 5415C centrifuge at 14000 rpm
for 5 min. The supernatant of each sample was transferred
into another centrifuge tube and evaporated to dryness under
a stream of nitrogen. The residues were reconstituted in
200 µl of 34% acetonitrile in water (initial mobile phase for
HPLC gradient). The sample solutions were filtered through
a Costar spin-X filter (2 µm) and transferred to autosampler
vials, and 20 µl of each solution were injected into the HPLC
system.
For the determination of protein-free etoposide and its
catechol metabolite, 500 µl human plasma samples were
filtered through Millipore (Bedford, MA, USA) Centrifree
filters and spun in a Beckman (Palo Alto, CA, USA) GS-
6KR centrifuge at 3920 rpm in a fixed-angle rotor (¾1900 g)
°
for 60 min at 10 C to remove protein and protein-bound
Cancer patient treatment and plasma sample
collection
etoposide and etoposide metabolites. The clear, colorless
filtrate was used for the quantitation of protein-free fractions
of etoposide and etoposide catechol. The procedure for the
protein-free fraction analysis was the same as for total plasma
except that 200 µl of protein-free plasma were used instead
of 50 µl of plasma and 800 µl of acetonitrile were used.
The studies were approved by the IRBs of the Children’s
Hospital of Philadelphia and the University of Pennsylvania.
In a typical clinical study, a pediatric cancer patient received
a 100 mg mꢀ2 dose of etoposide as a 1 h infusion daily
for 5 days. On days 1 and 5, blood samples were taken at
selected points before and after completion of the infusion.
The blood samples were kept on ice and centrifuged at
Calibration curves
°
The total and protein-free fractions of etoposide and
its catechol metabolite in human plasma were analyzed
separately. Calibration curves were prepared by adding
known amounts of etoposide and etoposide catechol to drug-
free human plasma at concentrations of 0.2–100 µg mlꢀ1
3000 rpm for 10 min at 10 C. Plasma was aliquoted and
°
stored in a freezer at ꢀ80 C immediately. To obtain protein-
free plasma, plasma (0.5 ml) was added to a Centrifree filter
°
(Millipore) and centrifuged at 10 C for 60 min at 1900 g. The
°
protein-free plasma samples were also stored at ꢀ80 C.
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781

Determination of etoposide and its catechol in plasma
775
fragmentation pathways of all three compounds are shown
in Fig. 3. In order to achieve the highest sensitivity for
selected reaction monitoring (SRM) analysis, the signal
intensities of the product ions from each compound were
optimized by changing the collision energy. For etoposide,
the transition of m/z 606 ! 229 was maximized with
a collision energy of ꢀ32 eV; for etoposide catechol, the
transition of m/z 592 ! 229 was maximized with a collision
energy of ꢀ25 eV; and for teniposide, the transition of
m/z 674 ! 383 was maximized with a collision energy
of ꢀ32 eV. The argon collision gas pressure was optimal at
3.5 m Torr for all three analytes.
Pharmacokinetic modeling
All modeling was conducted using the NIH-based Win-
SAAM kinetic modeling software.29 This uses a weighted
least-squares approach to fit the etoposide and catechol data
simultaneously, producing direct estimates of the adjustable
rate constants. For all data fitting, rate parameters were con-
sidered resolved when their fractional standard deviations
(FSDs) were <0.5 (or 50%). In fact, no estimated parameter
had an FSD >20%.
RESULTS AND DISCUSSION
Mass spectrometry
Full-scan mass spectra of etoposide, etoposide catechol and
teniposide showed abundant ions at m/z 606, 592, and
674, respectively. These ions corresponded to ammonium
adduct ions, [M C NH4]C, for each of the compounds.
Product ion spectra of [M C NH4] m/z 606 (etoposide),
m/z 592 (etoposide catechol) and m/z 674 (teniposide) at
a collision energy of ꢀ25 eV are shown in Fig. 2. Product
ions were formed through two major pathways: loss of
the sugar moiety and the phenolic side-chain. Proposed
Liquid chromatography
Separation of the analytes and internal standard was accom-
plished in 10 min using a gradient of acetonitrile–water
containing 5 mM ammonium formate and 0.1% formic acid.
Representative SRM chromatograms are shown for blank
human plasma spiked with the teniposide internal standard
[Fig. 4(a)] and blank protein-free plasma spiked with inter-
nal standard [Fig. 5(a)]. There were no interfering peaks at
the retention times for etoposide or etoposide catechol. No
interfering peaks were observed at the retention time of
the internal standard (data not shown). Representative chro-
matograms are also shown for human plasma [Fig. 4(b)] and
protein-free human plasma [Fig. 5(b)] spiked with etoposide,
etoposide catechol and internal standard.
Etoposide catechol eluted at 4.2 min, etoposide at 5.7 min
and the internal standard at 8.6 min. All three compounds
were resolved completely with acceptable peak shape. The
SRM response of teniposide was approximately twice that of
etoposide under the conditions of the assay. This resulted in
a peak area ratio of ¾0.5 when equal amounts were injected
on-column. The peak width for etoposide was significantly
greater than that for teniposide so that the peak height of
etoposide was considerably lower than that of teniposide
(Figs 4 and 5).
100
a
229
80
60
40
20
0
606
435
589
383
247
185
150 200 250 300 350 400 450 500 550 600 650
100
80
60
40
20
0
229
b
369
575
592
435
247
285
185
150 200 250 300 350 400 450 500 550 600 650
[MNH4]+
100
80
60
40
20
0
383
c
H+
674
657
H+
O
O
O
O
R2
229
O
O
O
HO
OH
O
503
O
O
O
H3CO
275
R1
O
OH
150 200 250 300350 400 450500 550 600650700
catechol m/z 369
etoposide m/z 383
teniposide m/z 383
catechol m/z 435
etoposide m/z 435
teniposide m/z 503
m/z
Figure 2. Product ion spectra of (a) etoposide, (b) etoposide
catechol and (c) teniposide (internal standard) at a collision
energy of ꢀ25 eV with argon at 3.5 mTorr. The parent ions of
etoposide, etoposide catechol and teniposide were at m/z 606,
592 and 674, respectively, owing to the formation of
ammonium adduct ions. A mobile phase of acetonitrile–
aqueous solution containing 5 mM ammonium formate and
0.1% formic acid was used.
OH
+
H+
H+
O
O
O
O
O
O
O
O
O
O
m/z 185
m/z 229
m/z 247
Figure 3. Proposed fragmentation pathways of etoposide,
etoposide catechol and teniposide.
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781

776
S. Pang et al.
100
50
a
Etoposide catechol
a
100
1.47E2
5.82E3
9.20E4
1.43E2
1.35E4
1.63E5
Etoposide catechol
50
0
0
100
100
Etoposide
Etoposide
50
50
0
100
0
100
8.60
8
Teniposide
8.56
8
Teniposide
50
0
50
0
3
3
4
5
6
7
9
10
4
5
6
7
9
10
b
100
50
Etoposide catechol
4.22
Etoposide catechol
1.47E2
100
4.23
b
1.43E2
50
0
100
5.72
0
100
Etoposide
5.82E3
5.73
Etoposide
1.35E4
50
50
0
8.58
8
100
0
1.18E5
10
Teniposide
7
8.56
8
100
1.46E5
10
Teniposide
50
0
50
0
3
4
5
6
9
3
4
5
6
7
9
Time (min)
Time (min)
Figure 5. Representative SRM chromatograms of (a) blank
protein-free plasma spiked with teniposide (0.5 µg mlꢀ1) and
(b) blank protein-free plasma spiked with etoposide catechol
(10 ng mlꢀ1), etoposide (100 ng mlꢀ1) and teniposide (internal
standard, 0.5 µg mlꢀ1). SRM conditions as in Fig. 4.
Figure 4. Representative SRM chromatograms of (a) blank
human plasma spiked with teniposide (internal standard,
0.5 µg mlꢀ1) and (b) blank human plasma spiked with
etoposide catechol (25 ng mlꢀ1), etoposide (0.5 µg mlꢀ1) and
teniposide (0.5 µg mlꢀ1). Positive ESI was performed in the
SRM mode with transitions of m/z 606 ! 229 for etoposide,
m/z 592 ! 229 for etoposide catechol and m/z 674 ! 383 for
teniposide (internal standard). Argon at 3.5 mTorr was used as
the collision gas. The collision energy was set at ꢀ32 eV for
both etoposide and teniposide and at ꢀ25 eV for etoposide
catechol.
(r2) were as follows: y D 0.0129 C 0.738x (r2 D 0.999)
for total etoposide (concentrations in µg mlꢀ1) and y D
0.00445 C 0.00375x (r2 D 0.998) for protein-free etoposide
(concentrations in ng mlꢀ1). A typical regression line for
total etoposide catechol (concentrations in ng mlꢀ1) was
y D 0.000342 C 0.000147x (r2 D 0.997) and for protein-
free etoposide catechol (concentrations in ng mlꢀ1) it was
y D 0.000593 C 0.000668x (r2 D 0.997).
Calibration curves
The calibration curves were plots of peak area ratios of
analyte/internal standard versus the analyte concentration.
The concentration of the standards ranged from 0.2 to
100 µg mlꢀ1 for total etoposide, from 25 to 15 000 ng mlꢀ1
for protein-free etoposide, from 10 to 5000 ng mlꢀ1 for
total etoposide catechol and from 2.5 to 1500 ng mlꢀ1
for protein-free etoposide catechol. The means, standard
deviations and RSDs were reproducible and demonstrated
linearity for all of the three days of the assay validation.
In order to achieve the best linearity, the data points were
fitted to a linear least-squares regression curve using the
weighting factors 1/x2, 1/x or a simple linear regression.
The weighting factor 1/x2 was found to provide the best fit
as determined by the r2 value and so this was used for all
curve fitting. The calibration curves were plotted with the
peak area ratio of the analyte to internal standard on the
y-axis and the concentration of the analyte on the x-axis.
Typical regression line equations and correction coefficients
Stability of etoposide catechol
Etoposide catechol is readily oxidized to the corresponding
quinone in the presence of oxidants, such as oxygen in
the air. It has been reported that ascorbic acid could
protect etoposide catechol from oxidation in the air.24 In
this study we examined the stability of etoposide catechol
at different concentrations over three days in the presence
of various concentrations of ascorbic acid. As shown in
Fig. 6, in the absence of ascorbic acid, etoposide catechol was
quickly oxidized to the quinone. In the presence of lower
concentrations of ascorbic acid (0.2 and 2 mM), etoposide
catechol was relatively stable, but only within 2–3 h. In the
presence of 50 mM ascorbic acid, both samples containing
10 and 1000 ng mlꢀ1 etoposide catechol showed a constant
area ratio for the three-day testing period. Therefore, 50 mM
ascorbic acid was added to the plasma samples to protect
etoposide catechol from oxidation. The stability of etoposide
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781

Determination of etoposide and its catechol in plasma
777
1200
1000
800
600
400
200
0
a
0.020
0.015
0.010
0.005
0.000
0
10
20
30
40
50
0
1
2
3
4
5
Time (h)
1.0
0.8
0.6
0.4
0.2
0.0
b
Figure 7. Etoposide catechol stability in plasma at room
temperature. Etoposide catechol was spiked at 1000 ng mlꢀ1
into duplicate plasma samples and left at room temperature.
Samples were taken at different time points and analyzed by
LC/MS. The horizontal line represents the starting
concentration of 1000 ng mlꢀ1
.
this study. The ascorbic acid also protected etoposide cate-
chol against adventitious oxidation during the work-up and
analysis procedure.
0
10
20
Time (h)
30
40
50
Precision and accuracy
The lower limits of quantification (LLQ) of total and
protein-free etoposide catechol were determined as 10 and
2.5 ng mlꢀ1, respectively. These were the lowest concentra-
tions that could be measured with a precision of š20%
and an accuracy between 80 and 120% (n D 5), the criteria
established by Shah et al.30 In order to quantify etoposide
and etoposide catechol simultaneously, the ratio of the two
analytes was set at 20 : 1 for total and 10 : 1 for the protein-
free fraction. Therefore, the LLQs for total and protein-free
etoposide were 200 and 25 ng mlꢀ1, respectively. The deter-
mined mean of LLQs of total etoposide and protein-free
etoposide were 0.216 µg mlꢀ1 and 27.07 ng mlꢀ1, respectively
(Table 1). For total and protein-free etoposide catechol the
LLQs were determined as 10.38 and 2.49 ng mlꢀ1, respec-
tively (Table 2). The intra-day precision of the LQC, MQC
and HQC ranged from 1.3 to 6.7% for total and protein-
free etoposide, respectively, and from 0.4 to 5.8% for total
and protein-free etoposide catechol, respectively (data not
shown). The inter-day precision for total and protein-free
etoposide for the LQC, MQC and HQC ranged from 0.7 to
3.2% and the accuracy (observed mean/theoretical ð 100)
was between 94 and 108% for all QC samples (Table 1). The
inter-day precision for total and protein-free etoposide cat-
echol for the LQC, MQC and HQC ranged from 1.0 to 5.2%
and the accuracy was better than 96–104% for all QC samples
(Table 2). The recommended30 intra- and inter-day precision
(RSD 15%) and accuracy (85–115%) requirements for the
LQC, MQC and HQC samples were met for all analytes.
Therefore, the method was sensitive, accurate and highly
reproducible for the determination of total and protein-free
etoposide and etoposide catechol in human plasma within a
wide dynamic range.
Figure 6. Etoposide catechol metabolite stability in the
presence of different concentrations of ascorbic acid as
antioxidant. Samples were prepared in duplicate with two
concentrations of etoposide catechol in plasma with different
concentration of ascorbic acid. (ž) 0.0 mM ascorbic acid; (
)
°
0.2 mM ascorbic acid; (ꢀ) 2 mM ascorbic acid; (ꢁ) 50 mM
ascorbic acid. (a) 10 and (b) 1000 ng mlꢀ1 etoposide catechol.
Area ratio is the peak area of the etoposide catechol to the
internal standard, teniposide.
catechol in the plasma at room temperature was also tested.
The plasma samples were left at room temperature and the
change in etoposide catechol concentration was monitored.
As shown in Fig. 7, etoposide catechol was stable for more
than 4 h, which was long enough to prepare the plasma
samples. The results also indicated that etoposide catechol
was stable in plasma and that no etoposide catechol was lost
during the blood sample treatment.
Reduction of etoposide quinone to catechol by
ascorbic acid
It has been reported that etoposide catechol readily under-
goes a two-electron oxidation to form etoposide quinone.3,4
Attempts were made to determine the quinone in the
presence of etoposide and etoposide catechol. However,
etoposide quinone proved to be too unstable for reliable
assay methodology to be developed. It was also reduced
to etoposide catechol in the electrospray ion source dur-
ing LC/ESI-MS analysis. Therefore, any quinone that was
present in the plasma was first reduced to the catechol using
ascorbic acid4 and it was not included as an analyte in
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781

778
S. Pang et al.
Table 1. Precision and accuracy for the determination of total
and protein-free etoposide in human plasma
(n D 3) and 4.4 and 12.5%, respectively, for plasma
ultrafiltrate (n D 3).
Compound
Parameter
LLQ
LQC
0.500
MQC HQC
Application of the method to pharmacokinetic
studies
Total etoposide Theoretical
0.200
5.00
50.0
(µg mlꢀ1
)
One representative pediatric cancer patient received a 100 mg
mꢀ2 dose of etoposide as a 1 h infusion daily for 5 days. On
days 1 and day 5, blood samples were taken at selected
time points before and after completion of the infusion.
The plasma was treated with ascorbic acid, which would
have converted any etoposide quinone back to the catechol.
Therefore, the catechol measurements would include any
quinone that may have been present at the time of sample
collection. The total and protein-free fraction of etoposide
and etoposide catechol in the plasma were determined by
LC/SRM/MS. A typical chromatogram of total etoposide
and total catechol from the patient 150 min after dosing is
shown in Fig. 8(a). A chromatogram showing protein-free
etoposide and catechol at the same time-point is shown in
Fig. 8(b).
In order to elucidate the mechanisms by which secondary
leukemias occur after etoposide treatment, it is important
to determine inter-patient differences in pharmacokinetic
parameters. This required the development of a rational
pharmacokinetic model. Using the WinSAAM modeling
program we obtained the best fit of the plasma concen-
tration–time parameters with the model shown in Fig. 9.
n
5
15
15
15
Mean
0.216
0.505
4.97
0.7
51.5
(µg mlꢀ1
RSD (%)
)
2.4
108
1.9
101
3.2
Accuracy
(%)
99.3
103
Free etoposide Theoretical 25.00 100.0
1000 10 000
(ng mlꢀ1
)
n
5
15
15
15
Mean
27.07 103.1
1006
9433
(ng mlꢀ1
RSD (%)
)
1.9
108
1.3
103
2.5
101
2.4
94.3
Accuracy
(%)
Table 2. Precision and accuracy for the determination of total
and protein-free etoposide catechol in human plasma
Compound
Parameter
LLQ
LQC MQC HQC
Total catechol Theoretical
10.00 25.00 250.0 2500
(ng mlꢀ1
)
a. Plasma (150 min)
n
5
15
15
15
1.73E3
4.23
100
Mean (ng mlꢀ1
RSD (%)
Accuracy (%)
)
)
10.38 24.87 249.1 2417
4.3
104
50
Etoposide catechol
Etoposide
1.2
99.5
3.5
99.7
5.2
96.7
0
100
5.74
1.73E6
2.03E5
Free catechol Theoretical
2.50 10.00 100.0 1000
(ng mlꢀ1
n
)
50
5
15
15
15
0
100
8.57
Mean (ng mlꢀ1
RSD (%)
Accuracy (%)
2.49 10.03 101.9 973
6.5
99.5 100
Teniposide
1.0
1.6
102
2.1
97.4
50
0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
b. Protein free plasma (150 min)
Sample stability and recovery
The processed standard samples were stable over 24 h when
100
50
4.32
1.86E2
°
left in the HPLC autosampler at 5 C (data not shown). In
m/z 592=>229@−32eV
order to determine the recovery, plasma and protein-free
plasma samples at MQC concentrations were processed as
described in the Experimental section and compared with
LC/MS analyses from the processed plasma blanks that
were spiked with standards at the MQC concentrations.
The recoveries for etoposide and etoposide catechol were
78.9 and 91.3%, respectively, from human plasma (n D 3)
and 99.0 and 101%, respectively, from protein-free plasma
(n D 3). To determine the plasma sample matrix effect on
analyte ionization, blank plasma and protein-free plasma
samples were processed as described in the Experimental
section, spiked at the MQC concentrations and compared
with LC/MS analyses of aqueous unextracted standards.
The suppression of ionization for etoposide and etoposide
catechol was 3.3 and 5.3%, respectively, for human plasma
0
100
5.88
9.57E4
50
m/z 606=>229@−25eV
0
100
8.61
9.89E4
m/z 674=>383@−32eV
50
0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
Time (min)
Figure 8. Representative SRM chromatograms of (a) plasma
and (b) protein-free plasma samples obtained from a pediatric
patient 150 min after receiving a 100 mg mꢀ2 dose of
etoposide as a 1 h infusion. SRM conditions as in Fig. 4.
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781

Determination of etoposide and its catechol in plasma
779
10
K21
K12
a
1
0.1
1
2
4
K01
K31
0.01
0.001
0.0001
K43
K34
3
0.00001
K03
1 = etoposide
3 = etoposide catechol
1
b
0.1
Figure 9. Pharmacokinetic model for etoposide metabolism to
its catechol metabolite. The k values represent the fractional
rates for transfer from one compartment to the next.
0.01
0.001
100
0.0001
0.00001
a
10
1
0
400
800
1200
1600
0.1
time [min]
0.01
0.001
0.0001
Figure 11. Plasma ultrafiltrate concentration–time profile for
subject A given a 100 mg mꢀ2 intravenous infusion of
etoposide over 1 h. Symbols as in Fig. 10. (a) Day 1; (b) day 5.
100
concentrations. The distribution spaces for etoposide and its
catechol metabolite were assumed to be the same, and the
inter-compartmental rate constants were assumed to have
the same value for the catechol movement. Unique values
of the etoposide inter-compartmental rate constants were
estimated from the data. To ease the fitting process, the
final (elimination) phase of the etoposide was fitted first,
and then the partition of that elimination between irre-
versible loss and conversion to the catechol was resolved
in subsequent data fitting, involving both etoposide and
its catechol metabolite. This made it possible for the first
time to determine simultaneous pharmacokinetic param-
eters for total and free parent drug and also its major
metabolite.
b
10
1
0.1
0.01
0.001
0.0001
0
400
800
1200
1600
time [min]
Figure 10. Plasma concentration–time profile for subject A
given a 100 mg mꢀ2 intravenous infusion of etoposide over
1 h. The solid triangles and solid squares are the plasma
concentrations determined by LC/MS for etoposide and its
catechol metabolite, respectively. The solid line represents the
calculated values for etoposide concentrations and the dashed
line represents the calculated values for the catechol
metabolite concentrations using the kinetic model in Fig. 9.
(a) Day 1; (b) day 5.
The plasma concentration–time curves for total etoposide
and etoposide catechol are presented in Fig. 10. On day
1, the peak plasma concentration of total etoposide (Cmax
)
was 24.02 µg mlꢀ1, which was achieved at the time point
(tmax) of 61 min, just after the end of the infusion. For total
etoposide catechol the Cmax was 106 ng mlꢀ1 and tmax was
90 min. The plasma concentration–time curves of protein-
free etoposide and etoposide catechol are shown in Fig. 11.
Cmax for protein-free etoposide and etoposide catechol on
day 1 were 1.16 and 11 ng mlꢀ1, respectively. On day 5,
both total and protein-free concentrations of etoposide were
similar to day 1 (Figs 10 and 11). However, the Cmax for
total (217 ng mlꢀ1) and protein-free (14 ng mlꢀ1) etoposide
catechol were significantly higher on day 5 when compared
with day 1 (106, 11 ng mlꢀ1) (Table 4). From our WinSAAM
This first-order two-compartment model has the same topo-
logy as that described by Relling et al.31 The model, which
takes into account both elimination and metabolism, was
able to fit the total plasma etoposide and etoposide cate-
chol as well as the free etoposide and etoposide catechol
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781

780
S. Pang et al.
Table 3. Pharmacokinetic parameters for total and protein-free etoposide from a patient dosed with 100 mg
mꢀ2 of etoposide on days 1 and 5
Total etoposide
day 1
Free etoposide
day 1
Total etoposide
day 5
Free etoposide
day 5
Parameter
Cmax (µg mlꢀ1
Volume of distribution (ml)
)
24.02
4086
41
1.16
102
49
249
145
198
279
23.37
5052
64
0.81
185
75
301
86
t
t
1/2˛ (min)
1/2ˇ (min)
AUC6h (µg mlꢀ1 min)
AUC24h (µg mlꢀ1 min)
Mean residence time (min)
255
237
3496
4957
295
3165
4043
247
109
249
Table 4. Pharmacokinetic parameters for total and protein-free etoposide catechol from a patient
dosed with 100 mg mꢀ2 of etoposide on days 1 and 5
Total catechol
day 1
Free catechol
day 1
Total catechol
day 5
Free catechol
day 5
Parameter
Cmax (ng mlꢀ1
)
106
24
193
25.0
56.8
11
35
217
24
205
56.7
93.0
14
24
ND
3.3
ND
t
t
1/2˛ (min)
1/2ˇ (min)
NDa
2.8
AUC6h (µg mlꢀ1 min)
AUC24h (µg mlꢀ1 min)
ND
a ND D not determined because concentrations were too low.
model (Fig. 9), the major pharmacokinetic parameters for
total and protein-free etoposide and etoposide catechol were
determined for an etoposide infusion of 100 mg mꢀ2 for 1 h
(Tables 3 and 4). The initial (distribution) half-lives (t1/2˛)
for total etoposide following infusion of etoposide were
41 and 64 min for days 1 and 5, respectively. This result
was in agreement with those reported previously.1,31,32 The
useful for optimizing etoposide dose and schedule so that
the risk of secondary leukemia is minimized.
CONCLUSIONS
We have developed and validated a sensitive and selective
LC/SRM/MS method for the determination of both total
and protein-free etoposide and its potentially toxic catechol
metabolite in human plasma. This method was able to quan-
tify etoposide and etoposide catechol simultaneously with a
concentration ratio up to 300 or higher in the plasma of pedi-
atric cancer patients. Good linearity, precision and accuracy
were achieved. The limit of quantitation was 10 ng mlꢀ1 for
total etoposide catechol and 2.5 ng mlꢀ1 for the protein-free
fraction. For total concentrations, simultaneous determina-
tion of etoposide and its catechol metabolite was possible
with only 50 µl of plasma. With simple sample treatment and
small sample requirements, this method has been applied
successfully to pharmacokinetic studies of pediatric can-
cer patient receiving intravenous etoposide. In addition, we
developed a pharmacokinetic model for a future pharmaco-
genetic study to determine if there is a correlation between
phenotype and genotype.15 The full pharmacokinetic stud-
ies will provide additional insight into the mechanisms by
which secondary leukemias may occur during etoposide
treatment. Finally, the LC/MS methodology may also make
it possible to optimize etoposide dose and schedule in order
to minimize the risk of leukemogenesis.
t
1/2˛ values for protein-free etoposide were similar to those
for total etoposide (Table 3). The t1/2˛ values for total and
protein-free etoposide catechol were slightly shorter than
those for the parent drug (Table 4). The terminal phase
elimination half-lives (t1/2ˇ) for total etoposide were 255
and 237 min for days 1 and 5, respectively, which was
within the range of 3.4–8.1 h reported previously.1,31,32
The t1/2ˇ values for total etoposide catechol were slightly
longer than those for the parent drug (Table 4). The
AUC24h for total etoposide was significantly lower on
day 5 compared with day 1 (Table 3). Interestingly, the
AUC24h (93 µg mlꢀ1 min) for total etoposide catechol on
day 5 was almost double that for day 1 (57 µg mlꢀ1min)
(Table 4). This trend was also observed for the protein-
free catechol, although the low protein-free concentrations
at 1440 min precluded a full pharmacokinetic analysis. This
suggested that the potentially toxic catechol metabolite could
accumulate during multiple-day dosing. Additional studies
in more patients will be required in order to confirm this
observation. However, it is noteworthy that a previous
study reported the accumulation of etoposide catechol
during etoposide therapy.33 Our pharmacokinetic model is
currently being used to correlate inter-patient differences
in etoposide and etoposide catechol pharmacokinetics with
CYP3A4 genotype.15 This information will ultimately be
Acknowledgements
We acknowledge the support of NIH grant CA77683 and support
from the NIH GCRC grant MO1-RR00240 and the Pediatric Phar-
macology Research Unit of the Children’s Hospital of Philadelphia
NIH grant UO1 5-U01-HD-37255.
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781

Determination of etoposide and its catechol in plasma
781
17. Strife RJ, Jardine I, Colvin M. J. Chromatogr. 1980; 182: 211.
18. Strife RJ, Jardine I, Colvin M. J. Chromatogr. 1981; 224: 168.
19. Manouilov KK, McGuire TR, Gordon BG, Gwilt PR. J. Chro-
matogr. B 1998; 707: 342.
20. Sinkule JA, Evans WE. J. Pharm. Sci. 1984; 73: 164.
21. Duncan GF, Farmen RH, Movahhed HS, Pittman KA. J. Chro-
matogr. 1986; 380: 357.
REFERENCES
1. Henwood JM, Brogden RN. Drugs 1990; 39: 438.
2. Relling MV, Nemec J, Schuetz EG, Schuetz JD, Gonzalez FJ,
Korzekwa KR. Mol. Pharmacol. 1994; 45: 352.
3. Van Maanen JM, Retel J, de Vries J, Pinedo HM. J. Natl. Cancer
Inst. 1988; 80: 1526.
4. Mans DR, Retel J, van Maanen JM, Lafleur MV, van Schaik MA,
Pinedo HM, Lankelma J. Br. J. Cancer 1990; 62: 54.
5. Demple B, Harrison L. Annu. Rev. Biochem. 1994; 63: 915.
6. Lee SH, Blair IA. Chem. Res. Toxicol. 2000; 13: 698.
7. Haim N, Nemec J, Roman J, Sinha BK. Biochem. Pharmacol. 1987;
36: 527.
22. Littlewood TJ, Hutchings AL, Bentley DP, Spragg BP. J. Chro-
matogr. 1984; 336: 434.
23. Cai X, Woo MH, Edick MJ, Relling MV. J. Chromatogr. B 1999;
728: 241.
24. Stremetzne S, Jaehde U, Schunack W. J. Chromatogr. B 1997; 703:
209.
8. Van Maanen JM, de Ruiter C, Kootstra PR, Lafler MV, de Vries J,
Retel J, Pinedo HM. Eur. J. Cancer Clin. Oncol. 1985; 21: 1215.
9. Van Maanen JM, de Vries J, Pappie D, van den Akker E,
Lafleur MV, Retel J, van der Greef J, Pinedo HM. Cancer Res.
1987; 47: 4658.
25. Liu B, Earl HM, Poole CJ, Dunn J, Kerr DJ. Cancer Chemother.
Pharmacol. 1995; 36: 506.
26. Stewart CF, Arbuck SG, Fleming RA, Evans WE. Clin. Pharmacol.
Ther. 1991; 50: 385.
27. Nemec J. US Pat. 4609644, 1986.
10. Van Maanen JM, Lafleur MV, Mans DR, van den Akker E, de
Ruiter C, Kootstra PR, Pappie D, de Vries J, Retel J, Pinedo HM.
Biochem. Pharmacol. 1988; 37: 3579.
28. Relling MV, Evans R, Dass C, Desiderio DM, Nemec J. J. Phar-
macol. Exp. Ther. 1992; 261: 491.
29. Greif P, ME, Wastney ME, Linares O, Boston R. Adv. Exper. Med.
Biol. 1998; 445: 3.
11. Osheroff N. Biochemistry 1989; 28: 6157.
12. Osheroff N, Zechiedrich EL, Gale KC. Bioessays 1991; 13: 269.
13. Walker PR, Smith C, Youdale T, Leblanc J, Whitfield JF, Siko-
rska M. Cancer Res. 1991; 51: 1078.
14. Li AP, Kaminski DL, Rasmussen A. Toxicology 1995; 104: 1.
15. Felix CA, Walker AH, Lange BJ, Williams TM, Winick NJ, Che-
ung NK, Lovett BD, Nowell PC, Blair IA, Rebbeck TR. Proc. Natl.
Acad. Sci. USA 1998; 95: 13176.
16. Lovett BD, Strumberg D, Blair IA, Pang S, Burden DA, Megoni-
gal MD, Rappaport EF, Rebbeck TR, Osheroff N, Pommier YG,
Felix CA. Biochemistry 2001; 40: 1159.
30. Shah VP, Midha KK, Dighe S, McGilveray IJ, Skelly JP,
Yacobi A, Layloff T, Viswanathan CT, Cook CE, McDowall RD,
Pittman KA, Spector S. Pharm. Res. 1992; 9: 588.
31. Relling MV, Yanishevski Y, Nemec J, Evans WE, Boyett JM,
Behm FG, Pui C-H. Leukemia 1998; 12: 346.
32. Crom WR, Glynn-Barnhart AM, Rodman JH, Teresi ME,
Kavanagh RE, Christensen ML, Relling MV, Evans WE. Clin.
Pharmacokinet. 1987; 12: 168.
33. Stremetzne S, Jaehde U, Kasper R, Beyer J, Siegert W, Schu-
nack W. Eur J. Cancer. 1997; 35: 978.
Copyright  2001 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2001; 36: 771–781