Journal of Analytical Toxicology, VoL 18, November/December 1994
Stability of Drugs of Abuse in Urine Samples Stored
S. Dugan and S. Bogema
American Medical Laboratories, 14225 Newbrook Drive, Chantilly, VA 22021
George Washington University School of Medicine, Washington, DC 20052
Department of Forensic Sciences, George Washington University, Washington, DC 20052
Isolated studies of the stability of individual drugs of abuse have
been reported. However, few have evaluated stability in frozen
urine samples stored for 12 months. We have determined the
stability of 11-nor-9-carboxy-Ag-tetrahydrocannabinol (9-COOH-
THC), amphetamine, methamphetamine, morphine, codeine,
cocaine, benzoylecgonine, and phencyclidine in 236 physiological
urine samples. Following the initial quantitative analysis, the
samples were stored at -20~ for 12 months and then reanalyzed.
All drug concentrations were determined by gas
chromatographic-mass spectrometric methods with cutoff
concenlrations of 5 ng/mt for 9-COOH-THC and phencyclidine
and 100 ng/mL for each of the other drugs. The average change in
the concentrations of these drugs following this long-term storage
was not exlensive except for an average change of -37% in
The Department of Health and Human Services issued
mandatory guidelines for federal workplace drug-testing pro-
grams in April, 1988 (1). These guidelines require that drug-
testing laboratories shall retain all confirmed drug-positive
urine samples for one year in frozen storage. During this year
of storage, retesting may be requested at any time. As a result
of storage, it is possible that the drug concentrations differ
from the initial results. Although storage studies of drugs have
been reported, generally they are restricted to a few drugs, are
short-term studies, or do not evaluate drug stability in frozen
urine. Therefore, this study was undertaken in order to eval-
uate the stability of drugs of abuse following 12 months of
Materials and Methods
Cocaine hydrochloride, codeine sulfate, d-amphetamine sul-
*Author to whom correspondence should be addressed.
fate, methamphetamine hydrochloride, morphine sulfate
pentahydrate, and phencyclidine hydrochloride were obtained
from The United States Pharmacopeial Convention (Rockville,
MD). Benzoylecgonine tetrahydrate was obtained from ADRI
(Park Forest, IL) and Sigma Chemical Co. (St. Louis, MO). 11-
nor-9-Carboxy-Ag-tetrahydrocannabinol (9-COOH-THC) was
obtained from Research Triangle Institute (Research Triangle
Park, NC). Morphine glucuronide was obtained from Supelco
Benzoylecgonine-d3, cocaine-d3, d-amphetamine-d3, and
methamphetamine-d5 were obtained from Sigma. 9-COOH-
THC-d3 was obtained from Research Triangle Institute. Phen.
cyclidine-d s was obtained from Radian Corp. (Austin, TX).
Codeine-d 3 and morphine-d3 were obtained from MSD Iso-
topes, Merck and Co. (Rahway, NJ).
Standards and solutions
A stock methanolic solution of 9-COOH-THC (100 lJg/mL)
was purchased from Research Triangle Institute. A stock solu-
tion of each of the other drugs (1 mg/mL) was prepared with
HPLC grade methanol and stored at -20~ Methanolic solu-
tions used as calibration standards were prepared at the fol-
lowing concentrations: amphetamine, methamphetamine, co-
caine, benzoylecgonine, codeine, and morphine, 1000 ng/mL;
9-COOH-THC, 80 ng/mL; and phencyclidine, 87 ng/mL.
Urine controls and blanks
DPC CON-DOA drugs-of-abuse urine controls, levels 1 and 3,
were obtained from Diagnostic Products Corp. (Los Angeles,
CA). We prepared blank negative urine control from samples
that had been analyzed and found to be drug free. National In-
stitute of Standards and Technology (NIST) 9-COOH-THC ref-
erence material was obtained from NIST (Gaithersburg, MD).
Cocaine reference material was obtained from the College of
American Pathologists (Chicago, IL), Both positive and nega-
tive quality-control urine samples were analyzed daily.
Chemicals and reagents
4-Carbethoxyhexafluorobutyryl chloride (4-CEHBC) was
purchased from PCR, Inc. (Gainesville, FL). N,O-bis-(Tri-
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission.
methylsilyl) trifluoroacetamide (BSTFA) with 1% trimethyl-
chlorosilane (TMCS) and tetramethylammonium hydroxide
(TMAH) were purchased from Pierce Chemical Co. (Rockford,
ILL Pentafluoropropionic anhydride (PFPA) was obtained from
Regis Chemical (Morton Grove, ILL Dichlorodimethylsilane
was purchased from Sigma. Iodomethane and N,N-dimethyl-
formamide dipropyl acetal (NNDDA) were purchased from
Aldrich Chemical Co. (Milwaukee, WI). [3-Glucuronidase, Type-
1, G 0251 (several different lot numbers were used), was ob-
tained from Sigma. HPLC grade acetonitrile, ethyl acetate,
hexane, methanol, and water were purchased from Fisher Sci-
entific (Fair Lawn, N J). All other reagents were American
Chemical Society grade chemicals and were purchased from
EMIT | drug-testing kits were obtained from the Syva Corp.
(Palo Alto, CA).
Chem Elut | and Analytichem Certify II | extraction columns
were obtained from Varian Associates (Sunnyvale, CA). "I~pe W
Prep | extraction columns were purchased from DuPont Co.
(Wilmington, DE). Clean Screen DAU SPE | extraction
columns were obtained from Worldwide Monitoring (Horsham,
PA). Jetube | extraction columns were obtained from HarLen
Medical (Gibsonia, PA).
Extractions were performed using the Vac Elut SPS 24
vacuum manifold from Varian or the Prep l obtained from
Electron-impact mass spectrometric analysis of urine ex-
tracts was performed on a 5890A Hewlett-Packard gas chro-
matograph interfaced to a 5970A Hewlett-Packard mass selec-
tive detector (MSD) obtained from Hewlett-Packard (Rockville,
MD). Separation was achieved with a Hewlett-Packard methyl
silicone cross-linked capillary column (12.5 m x 0.2-ram i.d.,
0.33-1Jm film thickness). Split injection was used, and helium
was the carrier gas. The MSD was operated in the selected ion
monitoring mode. A Hewlett-Packard printer, 7673A automatic
sampler, and color display monitor were used in conjunction
with ChemStation software (Version 3.2) for data analysis.
The 236 samples analyzed were obtained from urine speci-
mens submitted as part of urine drug-screening programs.
Presumptive analysis was conducted using the Syva EMIT
reagent system on a Hitachi 717 analyzer. Drug concentra-
tions were determined by gas chromatographic-mass spectro-
metric (GC-MS) methods using calibration standards specific
for the drug being analyzed. Results were determined on one
aliquot of the sample. Following the presumptive analysis,
positive samples were stored at 4~ generally for no longer
than 24 h prior to drug quantitation. Following confirmation
and quantitation of positive samples, the samples were trans-
ferred to frost-free frozen storage at -20~
stored in the same containers in which they were submitted for
initial testing. A variety of plastic containers were received.
After 12 months of storage, the drug concentrations were de-
Journal of Analytical Toxicology, Vol. 18, November/December 1994
Amphetamine and methamphetamine (2,3). Amphetamine
and methamphetamine were extracted from urine using Chem
Elut or Type W Prep extraction columns. A 2-mL sample,
buffered to pH 10.5, was applied to the column, following which
the drug was eluted with either four portions of 6 mL
hexane-isoamyt alcohol (99:1, Chem Elut) or 1.7 mL methylene
chloride-acetic anhydride (95:5, Type W Prep). The etuate from
the Type W Prep column was dried, the residue was reconsti-
tuted with 100 pL chloroform-isobutanol (95:5), and 2 pL of the
reconstituted sample was injected into the GC-MS. The GC in-
jection port temperature was 200~ The oven temperature,
initially held at 88~ for 0.1 min, was then increased to 290~
at a rate of 20~ The ions monitored were as follows: am-
phetamine, 44, 86, and 118; amphetamine-d3, 47; metham-
phetamine, 58, 100, and 191; and methamphetamine-d5, 62.
The eluate from the Chem Elut was collected and extracted
with 0.3N H2SO4, following which the acid layer was alkalinized
with 2N NaOH and extracted with n-butyl chloride. The sample
tube was immersed in a dry ice-isopropanol bath to freeze the
aqueous layer, and the n-butyl chloride layer was decanted. 4-
CEHBC (201JL) was added to the n-butyl chloride extract, and
the mixture was incubated at 6045~ for 30 min. Anhydrous
ethanol (300 vL) was added, and the sample was incubated
for another 30 min at 60-65~ Sample size was reduced to ap-
proximately 25 t~L, 100 ~L ethyl acetate was added, and 4 laL of
the mixture was injected into the GC-MS. The GC injection
port temperature was 170~ The oven temperature, initially
100~ was increased to 200~ at 30~
tored were as follows: amphetamine, 248, 266, and 294; am-
phetamine-da, 297; methamphetamine, 262, 280, and 308; and
Cocaine and benzoylecgonine (4,5). Cocaine and ben-
zoylecgonine were extracted from urine using either Clean
Screen DAN SPE extraction columns or Type W Prep columns.
When the Type W Prep columns were used, a 2-mL sample,
buffered to pH 9.5, was applied to the column, and the drugs
were eluted with deionized water and methylene chloride-
isopropanol, 9:1. The drugs were derivatized with 50 pL
NNDDA. Chloroform (100 IJL) was added prior to injection
into the GC-MS. When the Clean Screen DAU SPE extraction
columns were used, they were conditioned with 3 mL
methanol, 3 mL deionized water, and 1 mL phosphate buffer,
pH 5.5. A 2-mL sample, buffered to pH 5.5, was added to the
column. The column was then rinsed with 2 mL deionized
water, 2 mL 0.1N hydrochloric acid, and 3 mL methanol. The
drug was eluted with two portions of 3 mL methylene chlo-
ride-isopropyl alcohol (80:20) containing 2% ammonium hy-
droxide. After the extracts were evaporated under nitrogen,
they were reconstituted with 100 IJL ethyl acetate and 50 pL
BSTFA containing 1% TMCS. This mixture was incubated at
80~ for 30 min, and 1 IJL of the derivatized sample was in-
jected into the GC-MS. For both derivatives, the GC injection
port temperature was 250~ The oven was initially held at
190~ for 0.1 min, increased to 280~ at 18~
for 5 min. When the drugs were derivatized with NNDDA, the
following ions were monitored: cocaine, 182, 272, and 303;
cocaine-d3, 185; benzoylecgonine, 210, 272, and 331; and ben-
zoylecgonine-d3, 213. When the drugs were derivatized with
The ions moni-
Journal of Analytical Toxicology, Vol. 18, November/December 1994
BSTFA-TMCS, the following ions were monitored: cocaine,
182,272, and 303; cocaine-d3, 185; benzoylecgonine, 240, 346,
and 361; and benzoylecgonine-d3, 243.
Codeine and morphine (6, 7). Urine samples for codeine and
morphine were hydrolyzed by incubating a 2-mL sample with
20,000 units of ~-glucuronidase, buffered to pH 4.8, at 37~ for
12 h. After incubation, 1 mL carbonate buffer, pH 9.5, was
added, and the entire mixture was applied to a Type W Prep
column, following which the drugs were eluted with 15%
methanol and methylene chloride-isopropanol, 9:1. The
eluting solvent was evaporated under nitrogen, and 50 IJL
PFPA was added to the residue. The mixture was incubated at
50-60~ the solvent was evaporated, and the residue was re-
constituted with 100 ~L heptane-ethyl acetate, 1:1. The deriva-
tized sample (1 lJL) was injected into the GC-MS. The GC in-
jection port temperature was 250~ The oven was initially
held at 170~ for 0.1 min, increased to 280~ at 15~
held for 2 rain. The ions monitored were as follows: codeine,
282, 445, and 446; codeine-d3, 448; morphine, 414, 430, and
577; and morphine-d3, 580.
Phencyclidine (3,8-10). A 2-mL sample, buffered to pH 10.5,
was applied to a Jetube column, and the drugs were eluted with
three portions of 6 mL hexane-isoamyl alcohol, 99:1. The
eluate was extracted with 1 mL 0.1N H2SO4, which was then
alkalinized and extracted with 100 IJL chloroform-isobutanoi,
95:5. A 10-1JL aliquot of the chloroform-isobutanol extract
was injected into the GC-MS. The GC injection port tempera-
ture was 200~ The oven was held at 100~ for 0.1 rain and
then increased to 280~ at a rate of 20~
itored were as follows: phencyclidine, 186, 200, and 242; and
9-COOH-THC (11). Two methods were used for the quanti-
tation of 9-COOH-THC. In the first method, a 5-mL sample
was incubated for 30 rain at 50--80~ with I mL 10N KOH and
2 mL methanol, following which the pH of the sample was ad-
justed to less than 3 with concentrated HCl and phosphate
buffer. The sample was extracted with 3 mL hexane-ethyl
acetate (7:1) and was then collected and evaporated to dryness.
Eighty microliters of 1.25% TMAH in dimethyl sulfoxide was
added to the residue and allowed to incubate at 25~ for at
least 2 rain but no more than 5 rain. Then, 5 IJL iodomethane
was added and allowed to incubate at 25~ for at least 5 rain but
no more than 10 rain. The mixture was acidified by the addition
of 200 ~L 0.2N HC1 and was extracted with 80 ~L of isooctane.
The isooctane extract (10 IJL) was injected into the GC-MS.
In the second method, a 3-mL sample was incubated with
100 ~L 10N KOH at 70~ for 15 rain. After cooling, 0.3 mL
glacial acetic acid and I mL 0.1N HCl were added. The sample
was centrifuged for 5 rain, and the supernatant was applied to
an Analytichem Certify II extraction column that had been
preconditioned by washing with 2 mL methanol and 2 mL of a
0.1M sodium acetate buffer containing 5% methanol. The
sample was drawn through the column by vacuum. The
column was rinsed with 8 mL methanol-water (1:1) and then
2 mL ethyl acetate. Following each rinse the column was dried
with vacuum. The 9-COOH-THC was eluted with two portions
of 2 mL hexane-ethyl acetate--acetic acid (75:25:1) and was col-
lected and evaporated under nitrogen. The residue was recon-
The ions mon-
stituted with 50 ~L BSTFA containing 1% TMCS and incubated
for 15 rain at 90~ A 1-1JL aliquot of the derivatized sample
was injected into the GC-MS.
The GC injection port temperature for both of the methods
was 270~ The oven temperature was held at 170~ for 0.1
min, increased to 290~ at 20~
min. The ions monitored when the TMAH derivative was
formed were as follows: 9-COOH-THC, 313, 357, and 372; and
9-COOH-THC-d3, 316. The ions monitored when the
BSTFA-TMCS derivative was formed were as follows: 9-COOH-
THC, 371, 473, and 488; and 9-COOH-THC-d 3, 374.
Deuterated standards of each drug used as an internal stan-
dard were added to the urine sample prior to extraction. Iden-
tification of an analyte was based upon a comparison of the re-
tention time of the extracted drug with a drug standard. The
retention time of the analyte was within 2% of the drug stan-
dard for cocaine, benzoylecgonine, codeine, morphine, phen-
cyclidine, and 9-COOH-THC and within 1% of the drug stan-
dard for amphetamine and methamphetamine. A single ion
was used for quantitation. Quantitation was based upon the
peak area of a single drug ion compared with the peak area of
the corresponding ion of the deuterated internal standard.
Three ions were used to identify each drug quantitated. The
ion ratios of the confirming ions of each drug were required to
be within • of the same ion ratios of the calibration stan-
The limit of detection (LOD), limit of quantitation (LOQ),
and cutoff concentration were determined for each drug. The
LOD and LOQ were determined as +3 and +10 standard devi-
ations of the mean, respectively, of 10 assays of negative urine.
The cutoff concentration was established as greater than the
LOQ. Samples containing drugs at concentrations equal to or
greater than the cutoff concentration were identified as posi-
tive, and those containing drugs less than the cutoff concen-
tration were identified as negative.
and maintained for 2
Results and Discussion
Table I presents the LOD, LOQ, and standard deviation for
each of the drugs quantitated over the period of this study. The
cutoff concentrations used were 5 ng/mL for 9-COOH-THC
and phencyclidine and 100 ng/mL for each of the other drugs.
The effect of storage on the detectability of drugs is pre-
sented in Table II. Of the 374 drug analyses conducted on the
236 samples, there were 288 (77%) that were positive both
before and after storage, 72 (19%) that were negative both be-
fore and after storage, 4 (1%) that were positive prior to storage
and negative after storage, and 10 (3%) that were negative
prior to storage and positive after storage. In those samples
identified as negative either before or after storage, either
parent drug or metabolite was detected at concentrations below
Except where noted, the data in Tables III-V were obtained
from the 288 analyses in which both the initial and final drug
concentrations were greater than the cutoff.
The concentrations determined in these samples are pre-
Table I. Characteristics of the Quantitative Methods
(ng/mL) Drug SD* CVt
21 (5"~ 0) u
* LOD = Limit of detection.
t LOQ = Limit of quantitation.
* SD = Standard deviation.
w CV = Coefficient of variation.
ql The numbers in parentheses represent the concentrations at which the SDs were
Table II. Effect of Storage on the Detectability of Drugs
Drug n* +/+f _/_, +/j _/+11
* n = Number of analyses of 236 specimens.
+/+ = Number of analyses that were positive both before and after storage.
* ~ = Number of analyses that were negative both before and after storage.
+/- = Number of analyses that were positive prior to storage and negative after
~n _/+ = Number of analyses that were negative prior to storage and positive after
Journal of Analytical Toxicology, Vol. 18, November/December 1994
sented in Table III and demonstrate a wide range of drug con-
centrations of approximately 2 orders of magnitude for each of
the drugs evaluated.
Table IV presents the number of drug concentrations that in-
creased, decreased, or remained unchanged over the storage
period. These data demonstrate that for most of the drugs
there was no observable trend in concentration changes.
Therefore, generally, it is not possible to predict with a high de-
gree of reliability whether the drug concentration of a specific
sample will increase or decrease as a result of long-term frozen
storage. The one exception was phencyclidine, for which 45 of
50 (90%) concentrations decreased.
A summary of the individual concentration changes that re-
sulted from storage is presented in Table V. These data demon-
strate that the changes in drug concentrations varied widely fol-
lowing storage. The changes ranged from a maximum decrease
of 87%, observed with cocaine, to a maximum increase of 252%,
observed with codeine. The greatest average concentration de-
crease was 37% for cocaine, and the greatest average concen-
tration increase was 14% for codeine. The concentrations of am-
phetamine, methamphetamine, and 9-COOH-THC changed the
Table IV. Fluctuation of Concentrations in Stored Urine
Number of samples
concentration concentration concentration*
* Includes those samples that were negative both before and after storage.
Table III. Initial and Final Drug Concentrations
average Drug n* range SD t range SD
179 - 9179
181 - 130,000
232 - 18,656
118 - 18,380
7 - 776
* n = Number of samples that were positive before and after storage.
t SD = Standard deviation.
7 - 774
louma[ of Analytical Toxicology, VoL I 8, November/December 1994
Table V. Drug Concentration Changes in Stored Urine Samples
(-35 to 30%)*
(-56 to 73%)
Cocaine 8 -37
n ?9 % change <-50% -50 to -26% -25 to 0%
39 +14 1
20 -I 0
0 2 18 10 2
3 1 3 1 0
4 11 10 9
t 8 10 1
2 12 16 8
9 21 10 4
4 41 4 O
24 130 8t 32
(-68 to 63%)
42 +9 1
(-70 to 147%)
48 ~1 1
(-51 to 11%)
50 -14 I
51 to 75% >75%
Journal of Analytical Toxicology, Vol. 18, November/December 1994
2. R.J. Czarny and C.L. Hornbeck. Quantitation of metham-
phetamine and amphetamine in urine by capillary GC/MS. Part II.
Derivatization with 4-carbethoxyhexafluorobutyryl chloride. J.
Anal. Toxicol. 13:257-62 (1989).
3. R.K. Lantz and R.B. Eisenberg. Evaluation of the JET technique for
extracting drugs from urine. Clin. Chem. 24:821-24 (1978).
4. D.M. Chinn, D.J. Crouch, M.A. Peat, B.S. Fink[e, and T.A.
Jennison. Gas chromatography-chemical ionization mass spec-
trometry of cocaine and its metabolites in biological fluids. J.
Anal. Toxicol. 4:37-42 (I 980).
5. H.H. McCurdy. Quantitation of cocaine and benzoylecgonine
after Jetube | extraction and derivatization. J. Anal. Ibxicol. 4"
6. A.S. Christophersen, A. Biseth, B. Skuterud, and G. Gadeholt.
Identification of opiates in urine by capillary column gas chro-
matography of two different derivatives. J. Chromatogr. 422:
7. N.B. Wu Chen, M.I. Schaffer, R. Lin, and R.J. Stein. Simultaneous
quantitation of morphine and codeine in biological samples by
electron impact mass fragmentography. J. Anal. Toxicol. 6:
8. D.N. Bailey and J.J. Guba. Gas-chromatographic analysis for
phencyclidine in plasma, with use of a nitrogen detector. Clin.
Chem. 26:437-40 (1980).
9. D.N. Bailey, R.F. Shaw, and J.J. Guba. Phencyclidine abuse:
Plasma levels and clinical findings in casual users and in phen-
cyclidine-related deaths. J. AnaL Toxicol. 2:233-37 (1978).
I0. D. Freehold, P. Jones, G. Sebrosky, and I. Sunshine. Testing for
basic drugs in biological fluids by solvent extraction and dual cap-
illary GC/NPD. J. Anal. Toxicol. 10:10-14 (1986).
11. J.D. Whiting and W.W. Manders. Confirmation of a tetrahydro-
cannabinol metabolite in urine by gas chromatography. J. Anal.
Toxicol. 6:49-52 (1982).
12. R.W. Romberg and M.R. Post. Stability of THC-COOH and ben-
zoylecgonine in frozen urine sample. Presented at the 43rd An-
nual Meeting of the American Academy of Forensic Scientists,
Anaheim, California, February 18-23, 1991.
13. B. Levine and M.L. Smith. Stability of drugs of abuse in biological
specimens. Forensic Sci. Rev. 2:147-56 (1990).
Manuscript received April I0, 1992;
revision received March 14, 1994.