Journal of Analytical Toxicology,Vol. 30, May 2006
High-Performance Liquid ChromatographicMass Spectrometric Determination of
Methamphetamine and Amphetamine Enantiomers,
Desmethylselegilineand Selegiline,
in Hair Samplesof Long-TermMethamphetamine
Abusersor SelegilineUsers
Kenichi Nishida 1, Shinzi Itoh 1, Naohide Inoue2, Keiko Kudo3, and Noriaki Ikeda 3,*
1Criminal Investigation Laboratory, Oita PrefecturePolice Headquarters, Oita 870-8502, Japan; 2TherapeuticCenter of Parkinson
Disease, Asaki Hospital, Fukuoka 811-4312, Japan; and 3Departmentof ForensicPathologyand Sciences, GraduateSchool of
Medical Sciences, Kyushu University, Fukuoka 812-8582, Japan
Introduction
Abstract
We devised a highly sensitive method for simultaneously
determining methampbetamine (MA) and amphetamine (AP)
enantiomers, desmetbylselegiline (DMSG) and selegiline (SG), in
human hair using a derivatization technique and high-performance
liquid chromatography-electrospray ionization-mass spectrometry
(HPLC-ESI-MS). MA and AP enantiomers and DMSG were
effectively converted to trifluoroacetic acid (TFA) derivatives, and
the sensitivity of MA and DMSG increased five times over
compared with that of free bases. The TFA derivatives of each
compound were stable within one week in a stock solution of
methanol or for 24 h in the HPLC mobile phase (mixture of
methanol and ammonium formate buffer). Each compound was
well separated, and calibration curves were linear in the
concentration range 0.04-40 ng/mg for MA enantiomers, SG and
DMSG, and 0.2-40 ng/mg for AP enantiomers. The accuracy and
precision of the method were evaluated, and relative standard
deviations were within 7%. Our method was successfully applied
to hair samples obtained from long-term MA abusers and SG users.
(+)-MA and (+)-AP were detected from three MA abusers at
concentrations of 0.79-20.85 and 0.04-3.30 ng/mg, respectively.
On the other hand, (-)-MA, (-)-AP, DMSG, and SG were detected
in three SG users at concentrations of 2.48-9.05, 0.72-3.10,
0.12-0.59, and 0-0.04 ng/mg, respectively. Based on our obtained
data, discrimination of MA abusers from SG users was considered
to be possible by comparing optical isomers of MA and AP, the
existence of DMSG and/or SG, and the concentration ratio of AP
to MA in hair samples.
* Author to whom correspondence should be addressed.
E-maih [email protected].
232
Methamphetamine (MA) is a strong central nervous system
stimulant, the abuse of which is a social problem in Japan.
Selegiline (SG) [(-)-deprenyl] is a selective monoamine oxidase-type B inhibitor that has been used for the treatment of
patients with Parkinson's disease (1,2). Both drugs have a
common characteristic in that MAand amphetamine (AP) are
excreted in body fluids, such as urine, blood, etc. Main compounds excreted in urine are unchanged: MA and AP in MA
abusers and desmethylselegiline (DMSG), MA, and AM in SG
users. In SG users, the parent drug is not excreted into urine,
and DMSG is excreted at only 1% of the given dose and
rapidly disappears from urine. Two days from the intake of
SG, MA and AP are the only compounds in urine (3). To distinguish medical SG users from illicit MA abusers, urinary
analysis for the optical isomers of MA and AP has been carried
out because MA abusers in Japan take mainly the (+)-enantiomer, which is more active than the (-)-enantiomer,
whereas SG is metabolized in the body to (-)-enantiomer
(3,4). Many reports have been published for the chiral analysis
of MA and AP (5-10).
In forensic science, segmental hair analysis has been used to
prove habitual abuse and/or to knowthe history of abuse (11,12).
The general mechanism of drug incorporation into hair is one in
which drugs (in the blood stream) enter hair in growing cells at
the base of the hair follicle (13). Therefore, the existence of unchanged SG and DMSGin hair was expected,and DMSGwas detected in hair samples (14-16). However,the amount of drugs in
hair is only of a trace level (ng/mg hair order) even in addicts,
Reproduction(photocopying)of editorialcontentof thisjournal is prohibitedwithoutpublisher'spermission.
Journal of Analytical Toxicology, Vol. 30, May 2006
thus highly sensitive methods are required. We analyzed MA
and AP as trifluoroacetic acid (TFA) derivativesand pentafluoropropionic acid (PFP) derivatives by liquid chromatography-electrospary ionization-mass spectrometry (LC-ESI-MS)
with a semi-micro octadesyl silica (ODS) column. This acylation
technique increased the sensitivity of MAabout 5 times and was
successfullyapplied to a HA abuser's hair (17).
We present here a method which can determine MAand AP
enantiomers, DMSG and SG, simultaneously using a TFA
derivatizationtechnique and a semi-micro chiral column. Practical application of the method to hair samples of HA abusers
and SG users of Parkinson' disease patients is also presented.
traction step by an Extrelut | NT3 column at the same time as
internal standard (IS) was added.
Hair samples and sample preparation
Hair samples were voluntarily obtained with informed consent under the ethical guidelines prepared by the Ministry of
Health, Labor, and Welfare in Japan. The samples came from:
one man and two women, MAabusers arrested by Oita Prefecture Police; and one man and two women, SG users who had
been treated with anti-Parkinson therapy in the Asaki Hospital. Hair samples were collected directly from scalps for HA
users, and deciduous hair was used for SG users. Hair was cut
Experimental
Reagents
(+)-MAhydrochloride was purchased from Dainippon Pharmaceutical Company (Osaka, Japan), and (-)-MA hydrochloride, (+)-APsulfate, and (-)-AP sulfate were received from the
Ministry of Health, Labour, and Welfare.SG hydrochloride and
DMSGhydrochloride were provided by Fujimoto Pharmaceuticals (Osaka, Japan). An internal standard (IS), N-n-buthylaniline, purchased from Wako Pure Chemical Industries
(Osaka, Japan) was dissolved in 0.1M hydrochloric acid to give
a concentration of i l~g/mL.Trifluoroacetic anhydride (TFAA)
was of gas chromatography grade (Nakarai Tesque, Kyoto,
Japan), methanol was of high-performance liquid chromatography grade, and the other chemicals used were of analytical
reagent grade. The Extrelut | NT3 column was purchased from
Merck (Germany).
Standard stock solutions (1 I~g/mLeach) of MAenantiomers
and AP enantiomers, SG and DMSG, were prepared by dissolving each drug in 0.1M hydrochloric acid. This solution
was further diluted prior to use with distilled water to prepare
appropriate concentrations.
Control human hair was collected from 10 healthy volunteers in our laboratory. Spiked hair samples were prepared by
adding aliquots of standard stock solutions to the control hair.
The addition of standard solutions were done prior to the ex-
[M+H]+
~- ]
249
[]~l+lqH4~l-
B
188 [M+H]+
C
1 ~70
D
= +0 ~, ,~ .~ ~, +, +. =,li'+, ,+ .+ I+. ~, ,,+ ,. ,~ = ~,,.
m/z
Figure 1. Mass spectra obtained from (+)- and (-)-MA-TFA (A), (+)- and
(-)-AP-TFA (B), SG (C), DMSG-TFA (D), and IS-TFA(E) by directly injected
ESI-MS.
34 A5,6
•]
~]
~'1 m/z 249
232
~9.
'
m/z 249
1
J/
7
........
B
.....
2
3 4 5
. . . . . . . . . . . . . . . . . . .
~, . . . . .
~] m/z 270
, ,,, ,. . . . .
~.] m/z 270
Ret=.t~ time(n~)
6
A
Retentimtime(rain)
Figure2, Total and extracted ion chromatograms obtained from control hair (A) and spiked hair containing 10 ng/mg each drug (B). Peaks: 1, (-)-AP-TFA;2, (+)AP-TFA; 3, (-)-MA-TFA; 4, (+)-MA-TFA; 5, IS-TFA; 6, DMSG-TFA; and 7, SG.
233
Journal of Analytical Toxicology, Vol. 30, May 2006
LC-MS conditions
The apparatus used was a Waters Alliance LC-MS (ZMD)
system, which consisted of a 2690 Separations Module (Waters)
and Platform LCZ (Micromass) equippedwith an ESI interface
in the positive ionization mode. The analytical column was a
Chiral DRUG| (150 m m x 2-ram i.d., Siseido, Tokyo,Japan).
The mobile phase consisted of two solvents; solvent A was
2ram ammonium formate buffer (pH 5) in methanol, and solvent B was 2mM ammonium formate buffer (pH 5) in water.
The mobile phase was programmed isocratic with 40% of solvent A from 0 to 20 min and then a linear gradient of 40-95%
of solvent A from 20 to 75 min. The flow rate was 0.1 mL/min,
and the temperature was 25~ The ESI inlet conditions were
as follows: ion source temperature, 120~ capillary voltage,
+3.5 kV; cone voltages, 30 V for MA-TFAenantiomers and
IS-TFA, 15 V for AP-TFA enantiomers, and 20 V for SG and
DMSG-TFA; and multiplier voltage, 650 V. Selected ion monitoring (SIM) mode was used for quantification of each compound with m/z 246 for MA-TFAenantiomers and IS-TFA,
m/z 249 for AP-TFAenantiomers, m/z 188 for SG, and m/z 270
for DMSG-TFA.
into 1 cm long segments from the root. All strands were
washed by sonication five times each with 1% sodium laurylsulphate (SDS) and methanol, respectively,and dried at room
temperature. Five mg of dried hair were accurately weighed
and transferred into a test-tube. To ensure decontamination,
the final methanol washes were evaporated by rotary evaporator, and the residue was injected onto the LC-ESI-MS following TFAAtreatment as described later.
To the test tube, 1 mL of 2.5M sodium hydroxideand 0.1 mL
of diethyl ether were added, and the mixture was left to stand
for 3 to 4 h at room temperature. After complete dissolution of
the hair, 1 mL of 2M hydrochloric acid was added for pH adjustment, and 25 IJL of IS solution (N-n-butylaniline) was
added to the dissolvedsolution. The resultant mixture was applied to an Extrelut | NT3 column. After standing for 5 rain,
drugs were eluted with 15 mL of ethyl acetate. The eluent was
concentrated by rotary evaporator to approximately 50 IJL and
transferred into a micro-sample-tube.
50 IJLof TFAAwas addedto the tube and heated at 60~ for 30
rain, and the reaction mixture was evaporatedto dryness under
a stream of nitrogen. The residue was dissolved in 50 l~L of
methanol, and a 5-pL aliquot was injected onto the LC-ESI-MS.
Table I. Calibration Curves and Detection Limits Obtained by the Present
Method
Concentration
Range
Compound (ng/mg)
(+)-MA
(-)-MA
(+)-AP
(-)-AP
SG
DMSG
0,04-40
0,04-40
0.2-40
0.2-40
0,04-40
0,04-40
Regression
Equation
y=
,v=
y=
y=
,v =
,v =
Correlation
Coefficient
0.1301 x - 0,4266
0.1194x- 0.1782
0.0618x-0.0838
0.0651 x - 0,2096
0,4062x + 1,387
0.0856x + 0,0243
r2 =
r2 =
r2 =
r2 =
r2 =
r2 =
0,998
0,999
0,999
0,998
0,999
0,995
Detection
Limit
(ng/mg)
0.01
0.01
0.05
0.05
0.01
0.01
Table II. Within- and Between-Day Reproducibilities of Determination of MA
and AP Enantiomers, SG and D M S G
Concentration
Added
Compound
(+)-MA
(-)-MA
(+)-AP
(-)-AP
SG
DMSG
Between-day (n = 5)
Found
(ng/mg)
(ng/mg)
(%)
(ng/mg)
RSD
(%)
1
10
1
10
1
10
1
10
1
10
1
10
1.03
9,91
1.01
9.70
0.96
10.08
0.98
9.80
0.99
9.59
1.00
9.88
2.39
1.55
3.65
1.69
3.76
1.74
3.30
2.27
4.06
1.41
3.13
2.99
0.97
9.70
0.99
9.91
0.98
9.96
0.99
9.70
0.97
9.63
0.97
10.17
3,75
3.84
4,65
3.93
3,19
3.87
4.77
1.23
6,97
4.98
2.24
4,86
* Relativestandarddeviation.
234
Within-day (n = 5)
Found
RSD*
Validation
To know the stability of TFA derivatives
in the mobile phase solutions, spiked hair
samples containing 50 ng of each drug
were treated using the present method,
and the derivatized extract was dissolved
in 250 tJL of the mobile phase (40% solvent A and 60% solvent B) instead of
methanol. An aliquot (5 t~L) of the mixture was injected into LC-ESI-MS at intervals of 2 h, and peak area of each compound was measured for 24 h.
Calibration curves were obtained by
plotting the peak-area ratio of each drug
to IS versus the amount of each drug
added to control hair. Within- and
between-day reproducibilities of the proposed method were evaluated using control hair samples containing 1 or 10
ng/mg each drug. These samples were analyzed as previously described over a period of 5 days.
Results
To know the most suitable condition in
ESI-MS, standard samples (50 ng each)
treated by TFAAwere dissolvedin the mobile phase (40% solvent A and 60% solvent B) and were directlyinjected into the
ESI-MS. As shown in Figure 1, strong protonated moleculer ions ([M+H]+) were
obtained in (+)- and (-)-MA-TFA,
DMSG-TFA, SG, and IS-TFA at m/z 246,
Journal of Analytical Toxicology, Vol. 30, May 2006
270, 188, and 246 with 30-, 20-, and 30-V of cone voltages, respectively. However, (+)- and (-)-AP-TFAshowed weak protonated ions compared with ammonium adduct ion ([M+NH4]§
or fragment ion ([C6HsCH2CHCH3]§ m/z = 119) in the corn
voltage range of 10-40 V. Therefore, ammonium adduct ion of
m/z 249 was selected for the measurement of (+)- and (-)-APTFAwith 15 V of cone voltage, which increased the sensitivity
more than 10 times in SIM mode.
The strongest ions for all analytes were obtained with 2 to
3ram of ammonium formate buffer among 1-10mM in the
mobile phase, and the pH of the buffer did not affect the sensitivity and separation of each compound within the examined pH 3-7. Therefore, 2ram and pH 5 of the ammonium
formate buffer was chosen for the best sensitivity and protection of the column, respectively. TFA derivatives of each compound were stable for 24 h in the mobile phase solution (40%
solvent A and 60% solvent B).
Figure 2 illustrates typical total and extracted ion chromatograms of control hair sample and spiked hair sample containing 10 ng/mg each drug. Although DMSG-TFAand IS-TFA
gave the same retention time in total ion chromatogram, both
compounds were separated by the monitoring ions of m/z 270
for DMSG-TFAand m/z 246 for IS-TFA (Figure 2B). The other
analytes showed complete separation even in total ion chromatograms. There were no interfering peaks in any control
hair samples examined. The abuse of metylenedioxymethamphetamine (MDMA)has been increasing recently in Japan. So
MDMAwas examined by our method. TFAderivatives of optical
isomers of MDMAshowed a strong [M+H]§ at m/z 290 with 20
V of cone voltages. The peaks of (-)-enantiomer and (+)-enantiomer were observed at 56.4 and 60.0 min, respectively.Therefore, the peaks of MDMA-TFAenantiomers were discriminated
from those of the target analytes.
Table I summarizes the parameters of calibration curves
and detection limits at a signal-to-noise ratio of 3. Calibration
curve of each compound was linear over the wide range, which
can cover the practical concentrations in hair samples. The detection limits were between 0.01 and 0.05 ng/mg in the SIM
mode. Sensitivity of MA enantiomers in this method is about
two times higher than those of column switching LC-MS in determination for free base (20), and MAenantiomers and DMSG
are five times more sensitive than a result of determination for
free base in our laboratory using the same apparatus. However,
sensitivity of AP enatiomers is not increased compared with
that of column switching LC-MS and determination results of
free bases in our laboratory. The minimum amounts of corn-
.,.,c
b ~ m/~ 249
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. . . . . . . . . . . .
.
.
.
.
.
.
"I m/z 27O
........
..........
+oi:
.
.
.
.
~. ~ m/z ~ss
i
9
~m
'
t ~ (mm)
~
(,~.)
Figure 3. Total and extracted ion chromatograms obtained from the final methanol wash (A) and hair sample for MA abuser (abuserA) (B). Peaks: 2, (+)-AP-TFA;
4, (+)-MA-TFA; and 5, IS-TFA.
m/~ 249
m/z 249
t ...........................
~"] ,Wz ts8
~',1
+t,
.......
.
~ m / z ms
(~n)
.
.
7
'
~1
l~m~ ~
.
~time
(~)
Figure 4. Total and extracted ion chromatograms obtained from the final methanol wash (A) and hair sample for SG user (patient C) (B). Peaks: 1, (-)-AP-TFA;
3, (-)-MA-TFA; 5, IS-TFA; 6, DMSG-TFA; and 7, SG.
235
Journal of Analytical Toxicology, Vol. 30, May 2006
a mixture of 4raM ammonium formate
and methanol (50:50, v/v)as mobile phase
at a flow rate of 0.1 mL/min, But SG was
not eluted with this mobile phase within
Concentration*(ng/mg)
120 min. Complete separation of optical
(+)-MA
(-)-MA
(+)-AP
(-)-AP
SG
DMSG
isomers and elution of all analytes from a
column within 80 min was carried out
MA abuser
ND
using gradient program. [M+H]+ ions for
ND
Abuser A
5.67 + 0.26
ND*
0.55 + 0.05
ND
ND
MA,
AP, and DMSG were shifted to a
ND
Abuser B
0.79 + 0.04
ND
0.04 + 0.01
ND
ND
higher
mass range from m/z 150, 136, and
Abuser C
20.85 _+1.03
ND
3.30 + 0.22
ND
ND
174 to m/z 246, 232, and 270 by TFA
SG user
derivatization, respectively.As many backPatient A
ND
7.72 + 0.07
ND
1.87 + 0.38 0.04 + 0.01 o.59 +_O.Ol
ground peaks derived from the mobile
ND
Patient B
ND
2.48 + 0.03
ND
0.72 _+0.09
o.12 _+o.o2
phase appeared around the mass range of
Patient C
ND
9.05 + 0.12
ND
3.10 + 0.11 0.03 _+ 0.04 0.37 _+0.01
m/z 150, this shift contributed to an increased sensitivity of each drug.
* Mean + standard deviation (n = 3).
t Not detected.
External contamination on the surface
of the hair shouldbe removedbefore analysis. Nakahara (12) reported that external
pound required for taking full scan spectra (100 m/z scan in
MA
contamination
on
the surface of the hair could be removed
0.5 s) were 0.5 ng/mg for MA enantiomers, SG and DMSG,
thoroughly
by
washing
hairs with 0.1% SDS two or three
and 3 ng/mg for AP.
times.
Therefore,
we
washed
hair samples by sonication five
Within- and between-day reproducibilities of the proposed
times
each
with
SDS
and
methanol
after cutting them into 1method are presented in Table II. Satisfactory data (n = 5)
cm
segments.
From
the
results
obtained
by the analysis of the
with relative standard deviations of under 7% for all comfinal
methanol
wash,
we
knew
that
the
external
contamination
pounds were obtained.
on
the
surface
of
the
hair
could
be
removed
by
this method.
Hair samples obtained from MAabusers (abuserA-C) and SG
However,
it
is
also
reported
that
trace
amount
of
MA and AP
users (patient A-C) were analyzed using the present method.
would
be
detected
when
control
hair
was
soaked
in
more than
Abuser A had been taking 0.04 g MAonce or twice a week, and
20
IJg/mL
or
0.5
lJmol/dm
3
of
MA
or
AP
solution
for
24 h or
abuser B had taken 0.02 g MAsix times in a month. Abuser C
overnight
(12,18).
It
was
presumed
that
part
of
the
adsorbed
didn't remember the exact time of abuse and the amount of
drugs permeated into the hair shaft rather than being adMA. The patients A-C had been taking SG orally for more than
sorbed on the surface. Therefore, the external contamination
one year. Patient A took 2 or 4 tablets (containing 2.5 mg SG
should not be underestimated. When a hair sample is anahydrochloride) a day, and patient B and C took 2 tablets a day.
lyzed, it is important not only to detect the metabolite of the
Total and extracted ion chromatograms obtained from the
drug
but also to investigate the past behavior of a suspect.
final methanol wash and hair samples of abuser A or patient C
As
concentrations
of MAin hair samples varied significantly
are shown in Figures 3 and 4, respectively,and quantification
because
different
amounts
of MAwere taken by abusers bedata are summarized in Table III. The final methanol washes of
cause
of
drug
tolerance,
the
data obtained in this study were
hair samples in abuser A and patient C presented no peaks of all
evaluated
based
on
the
concentration
ratio of each compound.
target compounds on the chromatograms, and the methanol
The
concentration
ratios
of
AP
to
MA
(AP/MA)
in the hair samwashes in the other samples gave the same results. Hair samples
of
MA
abusers
and
SG
users
were
0.05-0.16
and 0.24-0.34,
ples of MAabusers A-C presented only (+)-MAand (+)-AP,and
respectively.
Nakahara
(19)
reported
that
the
ratio in MA
those of patient A-C gave (-)-MA, (-)-AP,DMSG,and SG. Conabusers'
hair
did
not
exceed
0.2.
The
ratio
in
our
experiment
centrations of analytes were 0.79-20.85 ng/mg for (+)-MAand
(0.05-0.16)
was
also
below
0.2
and
was
approximately
the same
0.04-3.30 ng/mg for (+)-AP in MA abusers' hair and were
value
that
had
been
reported
in
the
other
papers
(20-23).
The
2.48-9.05 ng/mg for (-)-MA, 0.72-3.10 ng/mg for (-)-AP,
ratios
in
SG
users'
hair
(0.24-0.34)
in
our
cases
were
almost
0.12-0.59 ng/mg for DNSG, and 0.03-0.04 ng/mg for SG in SG
same as those of Kronstrand eta]. (0.39), who studied longusers' hair. SG was not detected in the hair of patient B.
term SG users (16), but were lower than those of Kikura and
Nakahara (0.32-0.54), who reported on short-time SG users
(15 mga day during five days) (14).
Discussion
In this experiment, we proved that TFA derivatives usually
used for gas chromatography are also useful for LC-MS analBoth PFP and TFA derivatives were successfully applied for
ysis, and we established a highly sensitive method that can
analysis of MA in our previous study using an ODS column
determine MA enantiomers and AP enantiomers, DMSG and
(17). However,PFP derivatives of (+)-MAand (-)-MAwere not
SG, simultaneously in hair samples of MA and SG users. Hair
separated even by the chiral column used in this study; thus,
analyses showed differences in optical isomers of MA and AP
TFA derivatives were used in this study. TFA derivatives of opand in the values of AP/MAbetween two groups, and it also
tical isomers of MAand AP were completely separated by using
showed the existence of DMSG in SG users, which was not alTable III. Concentrations of MA and AP Enantiomers, SG and DMSG
Determined in Hair Samples by the Present Method
236
Journal of Analytical Toxicology,Vol. 30, May 2006
ways detected in urine samples. Thus, the present method
should be useful to distinguish therapeutic SG users from MA
abusers in forensic toxicological analysis.
Acknowledgement
We thank M. Ohara (Fukuoka, Japan) for language assistance.
References
1. W. grikmayer, P. Riederer, M.B.H. Youdim, and W. Linauer. The
potentiation of the anti-akinetic effect after L-dopa treatment by
an inhibitor of MAO-B, Deprenil. J. Neural Transm. 36" 303-326
(1975).
2. J. Knoll. The possible mechanisms of action of (-)-deprenyl in
Parkinson's disease. J. Neural Transm. 43:177-198 (1978).
3. M. Hasegawa, K. Matsubara, S. Fukushima, C. Maseda, T. Uezono, and K. Kimura. Stereoselective analyses of selegiline
metabolites: possible urinary markers for selegiline therapy.
Forensic Sci. Int. 101:95-106 (1999).
4. H.S. Shin. Metabolism of selegiline in humans. Drug Metab.
Dispos. 25:657-662 (1997).
5. M. Katagi, M. Nishikawa, M. Tatsuno, T. Miyazawa, H. Tsuchihashi, A. Suzuki, and O. Shirota. Direct analysis of methamphetamine and amphetamine enantiomers in human urine by
semi-microcolumn HPLC/electrospray ionization mass spectrometry. Jpn. J. Toxicol. Environ. Health 44:107-115 (1998).
6. T. Nagai and S. Kamiyama. Assay of the optical isomers of
methamphetamine and amphetamine in rat urine using high-performance liquid chromatography with chiral cellulose-based
columns. J. Chromatogr. B 525:203-209 (1990).
7. A. Suzuki, O. Shirota, Y. Ohtsu, Y. Makino, S. Ohta, and
T. Nagano. Direct analysis of methamphetamine enantiomers in
urine by semi-microcolumn liquid chromatography with photodiode array detector: application to forensic analysis. Chromatography 18:27-30 (1997).
8. Y. Makino, S. Ohta, and M. Hirobe. Enantiomeric separation of
amphetamine by high-performance liquid chromatography using
chiral crown ether-coated reversed-phase packing: application to
forensic analysis. Forensic Sci. Int. 78:65-70 (1993).
9. R. lio, S. Chinaka, S. Tanaka, N. Takayama, and K. Hayakawa.
Simultaneous chiral determination of methamphetamine and its
metabolites in urine by capillary electrphoresis-mass spectrom-
etry. Analyst 128:646-650 (2003).
10. E.M. Kim, H.S. Chung, K.J. Lee, and H.]. Kim. Determination of
enantiomeric metabolites of I-deprenyl, d-methamphetamine,
and racemic methamphetamine in urine by capillary electrophoresis: comparison of deprenyl use and methamphetamine
use. J. Anal. Toxicol. 24' 238-244 (2000).
11. T. Inoue and S. Seta. Analysis of drugs in unconventional samples.
Forensic 5ci. Rev. 4:90-107 (1992).
12. Y. Nakahara. Hair analysis for abused and therapeutic drugs.
J. Chromatogr. B 733:161-180 (1999).
13. G.L. Henderson. Mechanisms of drug incorporation into hair.
Forensic 5ci. Int. 63" 19-29 (1993).
14. R. Kikura and Y. Nakahara. Hair analysis for drugs of abuse. IX.
Comparison of deprenyl use and methamphetamine use by hair
analysis. Biol. Pharm. Bull. 18:267-272 (1995).
15. R. Kronstrand, M.C. Andersson, J. Ahlner, and G. Larson. Incorporation of selegiline metabolites into hair after oral selegiline intake. J. Anal. Toxicol. 25:594-601 (2001).
16. R. Kronstrand, J. Ahlner, N. Dizdar, and G. Larson. Quantitative
analysis of desmethylselegiline, methamphetamine and amphetamine in hair and plasma from Parkinson patients on longterm selegiline medication. J. Anal. Toxicol. 27:135-141 (2003).
17. K. Nishida, M. Narihara, K. Tsutsumi, and S. Ito. Analysis of
methamphetamine and amphetamine in human hair using acylated derivatives by LC/MS. Jpn. J. Sci. Techol. Iden. 6:109-115
(2002) in Japanese.
18. O.Y. AI-Dirbashi, N. Kuroda, S. lnuduka, F. Menichini, and
K. Nakashima. HPLC with fluorescence detection of methamphetamine and amphetamine in segmentally analyzed human
hair. Analyst 124:493-497 (1999).
19. Y. Nakahara. Detection and diagnostic interpretation of amphetamines in hair. Forensic Sci. Int. 70:135-153 (1995).
20. A. Miki, M. Katagi, and H. Tsuchihashi. Determination of
methamphetamine and its metabolites incorporated in hair by
column-switching liquid chromatography-mass spectrometry.
J. Anal. Toxicol. 27:95-102 (2003).
21. N. Takayama, S. Tanaka, and K. Hayakawa. Determination of
stimulants in a single human hair sample by high-performance
liquid chromatographic method with chemiluminescence detection. Biomed. Chromatogr. t l : 25-28 (1997).
22. J. Liu, K. Hara, S. Kashimura, M. Kashiwagi, and M. Kageura. New
method of derivatization and headspace solid-phase microextraction for gas chromatographic-mass spectrometric analysis of
amphetamines in hair. J. Chromatogr. B. 758:95-101 (2001).
23. K. Nakashima, A. Kaddoumi, Y. Ishida, T. Itoh, and K. Taki. Determination of methamphetamine and amphetamine in abusers'
plasma and hair samples with HPLC-FL. Biomed. Chromatogr. 17"
471-476 (2003).
Manuscript received May 17, 2005;
revision received December 13 2005.
237