Original Article Ann Clin Biochem 1996; 33: 344-346

A simple HPLC method for the separation of amphetamine

isomers in urine and its application in differentiating between
'street' amphetamine and prescribed D-amphetamine
S Palfrey and M Labib
From the Department of Clinical Biochemistry, Russells Hall Hospital, Dudley,
West Midlands DYl 2HQ, UK

SUMMARY. D-amphetamine has been increasingly prescribed to treat amphetamine

abusers. Prescribing D-amphetamine requires laboratory evidence or confirmation of
current use of 'street' amphetamine, using a method which should be capable of
differentiating between 'street' amphetamine and prescribed D-amphetamine. We
have developed a simple high-performance liquid chromagraphy (HPLC) method for
the separation of the two isomers of amphetamine in urine and have assessed its use
in differentiating between 'street' amphetamine and prescribed D-amphetamine. The
method is reproducible, free from interference and has a detection limit of 0·1 j1g(mL
for each isomer. Urine from patients prescribed D-amphetamine contained only a
trace amount of L-amphetamine (less than 4%) whereas urine from those taking
'street' amphetamine contained more than 50% L-amphetamine. The method is
applicable to confirmation of 'street' amphetamine misuse and for monitoring
patient compliance with treatment. The presence of 4% or less L-amphetamine in
urine would suggest that the patient is only taking prescribed D-amphetamine
whereas the presence of L-amphetamine in higher concentrations suggests that the
patient is taking 'street' amphetamine, with or without prescribed D-amphetamine.

Additional key phrases: drugs of abuse screening; L-amphetamine; chiral separation

In recent years, amphetamine has become the
second most widely used illicit drug in the UK,
after cannabis.' In many parts of the country,
amphetamines are apparently much more com-
monly used than opiates, with over 60% of
abusers injecting the drug. One pilot needle
exchange scheme in Portsmouth showed that the
majority (67%) of those attending the scheme
were amphetamine misusers.?
Prescribing a substitute drug, such as metha-
done, is a widely accepted intervention for
dependent opiate users." Although D-amphet-
amine prescribing is not yet widely employed for
amphetamine users, several recent trials have
reported positive findings and it is likely that D-
amphetamine prescribing will increase in the
future."
Amphetamine prescribing requires laboratory
evidence or confirmation of current use of
Correspondence: Mr S Palfrey.
'street' amphetamine. More importantly, the
analytical method employed should be capable
of differentiating between 'street' amphetamine
and prescribed D-amphetamine exists in two
forms; the L- and D-isomers. Illicitly-produced
or 'street' amphetamine sulphate is a mixture of
the two isomers whereas prescribed D-amphet-
amine sulphate is predominantly the D-isomer.
The chiral separation of the optical isomers of
amphetamine in biological fluids has been
described using HPLC6 and gas chromatogra-
phy-mass spectrometry (GC(MS)Y However,
these methods are complex or require special
equipment. We have developed a simple HPLC
method for the separation of the two isomers of
amphetamine in urine and have assessed its use
in differentiating between 'street' amphetamine
and 'prescribed' D-amphetamine. The method
described has been developed to complement
our existing procedure for drugs of abuse
screening using the Toxi-Lab system.

344
 Separation of amphetamine isomers

MATERIALS AND METHODS (a)

The HPLC system consisted of an SIL-6B
autosampler, an LC-6A pump and an LCD-
6AV detector set at 240 nm (Shimadzu, Dyson
Instruments Ltd, Houghton-Le-Spring, UK).
A Supe1cosi1 LC-(S)-naphthy1-urea column
25 em x 4·6 mm (Supelco UK, Poole, UK) was
used for separation of D- and L-amphetamine.
The mobile phase was hexane: propan-z-ol :
acetonitrile (97: 1 : 2) with a flow rate of 2·0 mL/
min. The column was cleaned after analysis by
pumping hexane: dichloromethane : methanol: -
propan-2-01(100: 37: 6: 40) for 10min.
The samples analysed were: (1) pure amphet-
amine isomers as 1mg/rnl, methanolic solutions (c) (d)
(Sigma Chemical Company Ltd, Poole, UK) 0·080
diluted (l: 100) with a drug-free urine; (2) a
III
dexamphetamine sulphate tablet (SmithKline c
Beecham Pharmaceuticals, Welwyn Garden "E
City, UK), which was crushed in water and tl
~
..c
'E a.
diluted to a concentration of 4 Jlg/mL; (3) 20 :::> E
urine samples from patients attending the 'drugs Ql "'!=
g 0·040 o
of abuse' clinic in Dudley, who tested positive <Il III

for amphetamine by Toxi-Lab; and (4) four -eo c

"E
~
Ul
urine samples from patients attending the 'drugs .0
« ..c
of abuse' clinic, who were prescribed dexam- a.
E
phetamine sulphate tablets. ~
Samples were prepared using Toxi A solvent
extraction tubes (Toxi-Lab, Microgen Ltd, I I I
Camber1ey,UK). Five millilitres of urine was added a 16 32 a 16 32

to a Toxi A tube and mixed by inversion for Retention time (min)

5 min, followed by centrifugation at 2000 rpm
for 5 min. At this stage, and before proceeding
with the Toxi-Lab protocol, 200 JlL of the
organic (top) layer was removed and put into a
micro autosamp1er vial. Before capping, 10 JlL
of 3,5 dinitrobenzoyl chloride in dichloro-
methane (0,1 g/25 mL) was added. Derivatiza-
tion is complete in about I min at room
temperature. Thirty microlitres of the deriva-
tized sample were then loaded onto the column.
The sample and dinitrobenzoyl chloride reagent
remain stable at 4°C for at least 4 weeks.
RESULTS
Using the conditions described, the extraction/
derivatization process produced two well-
defined peaks at approximately 24 and 28 min
corresponding to the L-amphetamine and D-
amphetamine, respectively (Fig. Ib). The limit of
detection (defined as the smallest quantity that
produced a peak distinguishable from baseline
noise) was 0·1 Jlg/mL for each isomer.
FIGURE 1. Chromatograms of (a) drug-free urine; (b)
standard containing 5 mg]L of both D- and L-
amphetamine in drug-free urine; (c) urine from a
patient on Dexedrine (D-amphetamine); (d) urinefrom
a patient who was attending the 'Drugs of Abuse' clinic
and who tested positive for amphetamines by Toxi-Lab.
All samples were prepared as described in the methods
section.
The Dexedrin solution produced two well-
defined peaks, a small peak corresponding to L-
amphetamine (-4%) and a larger peak that
eluted in the position of D-amphetamine
(- 96%; Fig. lc). A similar pattern was observed
in all four patients who were prescribed D-
amphetamine (Dexedrin). In none of these
patients did the L-amphetamine concentration
exceed 4% of the total concentration of
amphetamines.
In the 20 urine samples from patients
attending the 'drugs of abuse' clinic, which
tested positive for amphetamine by Toxi-Lab,
both isomers were found but the L-isomer was
always present in greater concentrations

Ann Clin Biochem 1996: 33

346 Palfrey and Labib
(53%-70%) than the D-isomer (Fig. ld). A third
peak eluting at approximately 3 min after the D-
amphetamine peak, but clearly separated from
it, was also seen occasionally.
The reproducibility of the assay was assessed
by analysing a urine spiked with a D-amphet-
amine solution (lOmg/L). The mean L-amphet-
amine content of this solution was measured at
3·9% (SD 0,15%) within batch (n= 10) and
3·8% (SD 0,29%) between batch (n = 10),
yielding within and between batch coefficient
of variation of 3·8% and 7,6%, respectively.
The chromatogram of a drug-free urine showed
no peaks after 18min (Fig. la), nor did any of the
following drugs: methamphetamine, methylene-
dioxyamphetamine (MDA), methylenedioxy-
methamphetamine (MDMA), N-ethylmethyl-
enedioxyamphetamine (MDE), methadone and
its metabolites, codeine, morphine, phenylpro-
panolamine and pseudoephedrine.
DISCUSSION
Prescription of D-amphetamine (Dexedrine
sulphate) by drug rehabilitation centres as a
substitute for 'street' amphetamine has created a
problem for the analyst, since normal screening
methods for amphetamines cannot differentiate
the D and L isomers. A positive screening result
will not distinguish whether the patient is taking
'street' amphetamine, or prescribed D-amphet-
amine, or both.
Separation of the optical isomers of amphet-
amine by chiral HPLC has been described
previously'-? but these methods were designed
for pharmaceutical preparations. Amphetamine
is first converted to the naphthoylamide deriva-
tive and then a column packed with 3,5-
dinitrobenzoylphenylglycine is used to resolve
the isomers." However, the separation isomers
was incomplete which made the measurement of
a small amount of one isomer, in the presence of
large amounts of the other, inaccurate. By using
a naphthyl-urea stationary phase and the 3,5-
dinitrobenzoyl derivative of amphetamine as
described here, complete separation of the
amphetamine isomers is achieved. An (R)-
naphthyl-urea column also resolves the isomers
but reverses their elution order compared with
the (S) column.
Although 'street' amphetamine is a racemic
mixture of the two isomers, L-amphetamine
predominates in urine because of the differing
rates of metabolism for D- and Lvisomers.'? Our
Ann cu« Biochem 1996: 33
finding that urine from all four patients
prescribed D-amphetamine contained only a
trace amount of L-amphetamine (less than 4%)
whereas urine from those taking 'street' amphet-
amine contained more than 50% of L-amphet-
amine, indicates that the method is capable of
identifying a 'street' amphetamine user.
CONCLUSION
This simple HPLC method could be used for
monitoring patient compliance with treatment.
Four percent L-amphetamine or less in urine
would suggest that the patient is only taking D-
amphetamine whereas the presence ofL-amphet-
amine in higher concentrations suggests that the
patient is taking racemic 'street' amphetamine,
with or without prescribed D-amphetamine.
REFERENCES
ISDD National Audit 0/ Drug Misuse in Britain.
London: Institute For The Study of Drug
Dependence, 1992
2 Stimson G. Injecting equipment exchange scheme
in England and Scotland. NIDA Res Monogr 1988;
80: 89-99
3 Cap1ehorn J, Bell J, Kleinbaum D, Gerbski V.
Methadone dose and heroin use during main-
tenance treatment. Addiction 1993; 88: 119-24
4 Fleming P, Roberts D. Is the prescription of
amphetamine justified as a harm reduction mea-
sure? J R Soc Health 1994; 114: 127-31
5 Cook B. Chirality of methy1amphetamine and
amphetamine from workplace urine samples. J
Anal Toxicol 1994; 18: 49-51
6 Zhou FX, Krull IS. Direct enantiomeric analysis of
amphetamine in plasma by simultaneous solid-
phase extraction and chiral derivatisation. Chro-
matographia 1993; 35: 153--9
7 Cody J, Scharzoff R. Interpretation of metham-
phetamine and amphetamine enantiomer data. J
Anal Toxicol 1993; 17: 321-6
8 Wainer W, Doyle T, Adams W. Liquid chromato-
graphy chira1 stationary phases in pharmaceutical
analysis: Determination of trace amounts of the
( - )-enantiomer in ( + )-amphetamine. J Pharm Sci
1984; 73: 1162-4
9 Pirkle W, House D, Finn J. Broad spectrum
resolution of optical isomers using chiral high-
performance liquid chromatographic bonded
phases. J Chromatogr 1980; 192: 143--58
10 Matin S, Wan S, Knight J. Quantitative determina-
tion of enantiomeric compounds: l-Simultaneous
measurement of the optical isomers of amphet-
amine in human plasma and saliva using chemical
ionization mass spectrometry. Biomed Mass Spec
1977; 4: 118-21
Accepted/or publication 29 August 1995