Download Technique for Rapidly Separating Drugs from

CLIN.
CHEM.
20/2,
307-309
(1974)
Technique for Rapidly Separating Drugs from Biological Samples
Vijay Aggarwal, Raymond Bath, and Irving Sunshine
With use of a sample/solvent
ratio of 8.5 ml to 100
zl, certain drugs may be extracted
efficiently
from
urine that has been saturated with a solid buffer. An
aliquot of this extract may be injected
gas chromatograph.
directly into a
is
of some drugs of abuse.
We show how this procedure
applied to confirmation
Additional
urine
Keyphrases:
#{149}
toxicology
gas chromatography
drug abuse
drugs in
#{149}
#{149}
Many commonly
used techniques
for detecting
drugs in
biological
samples
(e.g., gas chromatography,
thin-layer
chromatography,
fluorimetry,
and ultraviolet
spectrophotometry)
require
that the drugs first be separated
from
the biological
matrix.
They usually
are separated
by extraction with an immiscible
organic solvent or by adsorp-
most other procedures,
in which relatively large volumes
of ether or chlorinated
hydrocarbons
are used as extractants. However, total sample preparation
time is still
long, because it is necessary to “clean up” the extract.
We have developed a faster and simpler procedure,
in
which the need for purification
of the initial extract is
eliminated without, in most cases, a concomitant
decrease
in the quality of the final extract. Most of our experience
with this technique, and the major concern of this note, is
with the separation
of organic bases from urine samples,
but the same principles can be applied to the separation
of barbiturates.
Materials and Methods
Reagents
tion onto a solid support.
These are time-consuming
methods, and frequently yield extracts that contain many
normally present substances
that may interfere with the
subsequent identification
procedure.
In an effort to avoid these problems, several investigatom have used solid buffers and small volumes of solvent
to extract drugs rapidly. Bastos et al. (1) reported that or-
All solvents and chemicals are ACS reagent grade.
Solid buffer: a 3:2 (by wt) mixture
of NaHCO3
K2C03.
ganic bases can be isolated
from urine by saturating
the
urine with potassium
carbonate
and then extracting
with
1.0 ml of ethanol.
This ethanol
extract
is then purified
by
shaking
it with a pH 8.5 buffer and ethyl ether. The drugs
in question
concentrate
in the ether-ethanol
phase. Horning et al. (2) reported
a separation
procedure
in which diluted plasma
is saturated
with potassium
carbonate
and
then extracted
with two 1.0-ml portions
of isopropanol.
This isopropanol
extract
is evaporated,
and the residue
dissolved
in methanol,
which is then washed
with isooctane. The isooctane,
which now contains
any lipid material, is discarded
and the washed methanol
extract
is used
for the analysis.
Use of small volumes of ethanol in these procedures
decreases
the solvent
evaporation
time as compared
with
Procedure
County Coroner’s Laboratory,
and the Department
School of Medicine, Case Western Reserve University, Cleveland, Ohio 44106.
Cuyahoga
of Pathology,
Received
Oct. 25, 1973;
accepted
Nov. 19, 1973.
Internal
standard:
5 mg
of
methoxyphenamine
and
in
100
ml of water.
Extraction
solvent:
a 4:1 (by vol) mixture
of chloroform
and isopropanol.
Add enough solid buffer (about 5 g) to saturate 10 ml of
urine. The resulting solution has a pH of 9.5. Shake the
mixture
well and centrifuge, to separate excess buffer and
any other insoluble material. Transfer 8 ml of the supernatant buffered urine to a 20-mi test tube that tapers to a
sharp tip designed to contain 10-20 zl (available
from
Fuchs and Sons, Pittsburgh,
Pa. or from Fisher Scientific,
711 Forbes Ave., Pittsburgh,
Pa. 15219). Add 0.5 ml of the
internal standard solution and, while mixing with use of a
vortex-type
mixer, add 100 zl of the extraction
solvent.
Vortex-mix the contents of the tube for 20 s after the solvent is added, then centrifuge for 5 mm. Aspirate and discard most of the supernatant
urine and with a 100-zl syringe, remove an aliquot of the solvent layer (usually
about 60 ,il). Use of the test tube with a sharp taper best
facilitates removal of the solvent layer, but if these specific tubes are unavailable,
conventional
tapered centrifuge
tubes may be used.
CLINICAL
CHEMISTRY,
Vol. 20, No.2,
1974
307
r
IS
Results and Discussion
Elution
Al
I
times
for the
acetyl
derivatives
of amphet-
amine,
methamphetamine,
and methoxyphenamine
are
close to those of the other organic bases; use of acetic anhydride
permits
a higher
initial
temperature
and thus
shortens
the time required
for gas chromatography.
Codeine is not completely
derivatized under the given conditions and so gives peaks for both codeine and acetylco-
a
Ii
deine.
The
principal
advantage
of this extraction
scheme is its
part to the use of the very small volof extraction
solvent.
Because
the volume
of the
speed, owing in large
umes
final extract
is so small,
it can be quickly
concentrated
at
room temperature,
obviating
the need for heating
baths,
air streams,
and the like. Complete
evaporation
must be
54
54
IS
IS
54
TISE IN MINUTES
Fig. 1. A: Analysisof a typicaldrug-free
urine (IS, internal standard).
B: Analysis of the same urine as in A, with
drugs added to a final concentration of 2.0 zg/ml of each
drug. C: Analysis of a urine sample from a methadonemaintained
patient.
Note
6) and acetylcodeine
were also observed
peaks
(peak
similar
to codeine
(peak
7). These same two peaks
when acetic
anhydride was omitted
3, caffeine;4,methadone metabolite;
5, methadone; 6,codeine;7,acetylcodeine;
8. quinine
1, amphetamine; 2, methamphetamine;
Table 1. Sensitivity and Extraction Data
Percent
cxtracted
Amphetamine
Methamphetamine
Methadone metabolite
Methadone
Codeine
Quinine
Based
#{176}
on
analysis
of
Senslvity,
CV, %(R)#{176} pg/mt
30
60
55
50
60
60
to which the
of 2.5 ,hg/mI. (n
of drug standard.
4.4
6.4
5.5
6 1
114
50
urines
added to a concentration
R = area of drug peak/area
were
=
respective
10).
0.5
0.1
0.3
0 5
05
01
drugs
avoided to ensure adequate recovery of volatile bases.
Figure 1A shows the gas chromatographic
trace obtained on analysis of a typical drug-free ,urine. Figure lB
shows the trace for the same urine after drugs and drug
metabolites
were added to it. There are no major unknown peaks in these two traces, which demonstrates
the
selective isolation of drugs from the biological sample
without the need for a “cleanup”
step. In addition,
as
shown
parable
in Table
to that
1, the amount
of drugs
achieved
by techniques
volumes of solvent are used.
Analyses of over 1000 urine samples
extracted
is comin which larger
obtained
values for methadone
and its metabolite were obtained by
a procedure similar to that just described.
When analyzed
as described above, about 80% of the
urine samples from patients
in methadone
maintenance
programs
yield peaks with retention
times identical to
those of codeine and acetylcodeine
(Figure 1C). They also
yield these same two peaks when acetic anhydride
is
omitted. This behavior is unlike that observed with authentic samples of codeine, so such urine samples presumably do not contain
codeine.
Because
of this problem, another analytical
be used if codeine has to be confirmed.
The extract can be analyzed by any of several different
techniques. After evaporation,
an aliquot of the residue may
be injected
into a gas chromatograph
without further
treatment
or after reaction with any one of several suitable derivatizing
reagents
such as N,O-bis(trimethylsilyl)acetamide,
trimethylsilylimidazole,
or acetic anhydride.
We routinely transfered 40 ,l of the extract into the small
aluminum
capsule used in the Perkin-Elmer
As-41 Autosampler. This capsule was allowed to stand at room temperature for 15 mm, to permit
most of the solvent
to
evaporate.
Before the capsule was sealed, 3 d of acetic
anhydride was added. The capsule was then injected into
a Perkin-Elmer
Model 900 gas chromatograph,
and chromatographed under the following conditions:
Column: glass, 3 ft x 1/4 inch (o.d.), packed with “3%
OV-17 on Chromasorb
W, 80/100
mesh” (Supelco, Inc.,
Bellefonte, Pa. 16823).
from pa-
tients
on a methadone
rehabilitation
program
produced
some peaks
with retention
times
identical
to those of
peaks 1 to 8 in Figure 1. The areas of these peaks were always smaller
than those used in calculating
the minimal
detectable
amounts
(sensitivity)
listed in Table
1. The
procedure
should
However,
all other urines can be analyzed
for codeine by this procedure, and conversely,
the absence
of peaks with retention
times of 15.5 and 16.3 mm. reliably
indicates
that codeine
is absent.
Morphine
also cannot
be determined
by this
procedure.
We used solid
NaOH or NH4OH
potassium
to adjust
carbonate
because use of
the pH of the urine samples
resulted
in the formation
of emulsions,
a tendency
decreased
by use of the solid buffer. An extraction
solvent
that is denser than sater
must be used, because
recovery
of the solvent
in the tapered
tip after centrifugation
is
much easier than attempting
to recover such a small volume from the surface
should
a low-density
solvent
be
used.
If the detection of amphetamine
in urine is of primary
concern, sensitivity may be improved by substituting
3 g
of Na2CO3 for the recommended
solid buffer. Although
recovery of the other bases is decreased,
the recovery of
Oven temp: 130 #{176}C
for 3 mm, then increased
by 12 #{176}C amphetamine
is about 60%.
per minute to 280 #{176}C,
and maintained
at 280 #{176}C
for 8 mm.
When barbiturates
are to be extracted from urine, a 3:1
Injection-port
temp: 250 #{176}C.
(by wt) mixture of NaH2PO4 and Na2HPO4 in place of
the NaHCO3 and K2C03 buffer can be used. Plasma, diManifold temp: 300#{176}C.
Gas flow rates: helium, 30 ml/min;
hydrogen, 30 ml!
luted with an equal volume of water, may also be used as
mm; air, 300 mi/mm.
the sample in this manner.
308
CLINICAL
CHEMISTRY,
Vol. 20, No.2,
1974
If the extract, prepared as described, is too contaminated for reliable analysis, a double small-volume
extraction
can be used. In this modification,
the buffered
urine is extracted with 1.0 ml of extraction solvent. After centrifugation, an aliquot of this extract is transferred
to a tube
containing 100 il of HC1 (0.4 mol/liter).
After the mixture
has been shaken
and centrifuged,
the HC1 layer is removed and made alkaline
with 50 jl of NaOH (1 mol/
liter).
This
tion solvent,
matograph.
We have
mixture
is then
an aliquot
used
the
extracted
with
50 il of extrac-
of which is injected
described
procedure
into the chroeffectively
CHEM.
20/2,
309-311
(b) Sample manipulation
is minimized, since a purification
of the initial extract is not
generally required. (c) Most importantly,
this procedure’s
brevity does not affect the quality of the extract, and
highly reliable information may be gained from its use.
Supported
in part
by grant
09863-11
the NIGMS,
from
NIH,
USPHS.
References
1. Bastos,
and
reliably for longer than a year as a confirmatory
test for
organic bases, other than morphine and codeine, in urine.
We favor the described procedure over other techniques
for several reasons. (a) Evaporation
of large volumes of
CLIN.
solvents is eliminated.
M. L., Kannanen,
G. L., Young, R. M., Montforte,
J.,
and Sunshine, I., Detection of organic bases and their metabolites
in urine. Clin. Chem. 16,931 (1970).
2. Horning, M. E., Boucher, E. A., Stafford, M., and Horning, E.
C., A rapid procedure for the isolation of dregs and drug metabolites from urine. Clin. Chim. Acta 37, 381 (1972).
(1974)
Single-pH Extraction Procedure for Detecting Drugs of Abuse1
Ramon E. Stoner and Connie Parker2
A method
is presented
for determining
certain drugs
of urine. The urine samto pH 6.0 with a buffer containing
and determined
phy (TLC).
of abuse in a single extract
ple is adjusted
bromcresol
by (e.g.)
use of thin-layer
chromatogra-
Materials and Methods
purple. Basic drugs, such as amphetam-
ines and narcotics,
form an organic salt with the ionized bromcresol
purple, which is extractable
with a
mixture of chloroform
and 2-propanol
(3:1 by volume). At pH 6.0, weak acids such as barbiturates
and neutral drugs such as glutethimide
are also soluble in this solvent. Consequently,
the major classes
of drugs are extracted
simultaneously.
The extract is
then concentrated
and the individual
drugs are determined
by thin-layer
chromatography
in a solvent
that will separate bromcresol
purple from the drugs.
Materials
Silber-Porter
50-mi centrifuge
H. Thomas,
Philadelphia,
Pa.
2619-F20, respectively).
tubes
19105;
(A.
with stoppers
No. 2619-B25
TLC development
apparatus (No. 6071; Eastman
N.Y. 14650).
TLC sheets, Eastman no. 6061 silica gel.
and
Kodak
Co., Rochester,
Micropipets,
10-id
(Camag
Inc.,
New
Berlin,
Wis.
53151).
Ultraviolet
view
box, containing
and 366 nm.
Spectrofluorometer,
halla, N.Y. 10595).
lamps
emitting
light
at
254
Additional
Keyphrases:
drug screening
thin-layer chromatography
We
would
cedure
needed
a means
save time without
#{149}toxicology
#{149}
screening
of screening
sacrificing
#{149}
procedure
urine
samples
proficiency.
The
that
pro-
described here was developed to extract acid, neutral, and basic drugs from urine at pH 6, with a buffer
containing
bromcresol purple. At this pH, acid and neutral
drugs
this
drugs
form
are soluble
an organic
in the organic
solvent.
The
salt with bromcresol
purple,
basic
and
salt (but not bromcresol
purple) is also soluble in the
solvent. Consequently,
the major classes of drugs are extracted simultaneously,
and may then be concentrated
Clinical
Laboratory
Service,
West
Side VA Hosp.,
820 S.
Damen Ave., Chicago, Ill. 60612.
1 Presented
in part at the 25th National
Meeting,
AACC, New
York, N. Y., July 15-20, 1973.
2 Present address: Dept. of Medicine,
University
of Illinois College of Medicine,
Chicago, Ill.
Received Oct. 13, 1973; accepted Nov. 26, 1973.
MK I (Farrand
Optical
Co., Val-
Gas chrornatograph,
(Hewlett
Packard
No. 5700A) fitted with 6 ft. x ‘ in. stainless-steel
column packed with
“3% OV 17 on HP Chromosorb
W” (Supelco,
Inc., Bellefonte, Pa. 16823)
Evaporating
bath, To handle a large number of samples
per day, one needs a method
of evaporating
the solvent
quickly.
Such a device may be constructed
by building
an
aluminum
tube rack to fit into a water bath (set at 70
#{176}C).
A 12-unit
hematocrit
tube cleaner
is mounted
over
this rack and a latex rubber
tube, 3 mm (i.d.) by 30 cm, is
connected
to each U-tube
of the cleaner.
To the other end
of each of the rubber tubes is connected
a 5-in. “Pharmaseal”
connector
(Scientific
Products,
McGaw
Park,
Ill.
60085; No. T5300). The free end of this connector is loosely inserted
into the sample
tube. The outlet of the hematocrit tube cleaner is connected
to a vacuum
line. Such a
unit produces
tubes.
a decreased
CLINICAL
pressure
CHEMISTRY,
that is the same for all
Vol.20,
No. 2, 1974
309