Nor-Ag-Tetrahydrocannabinol-9 - Journal of Analytical Toxicology

Journal of Analytical Toxicology, Vol. 28, January/February 2004
An Investigation of the Stability of Free and
Glucuronidated -Nor-Ag-Tetrahydrocannabinol-9 -
carboxyl,c Acid lAuthentic Urine Samples
Gisela Skopp 1,* and Lucia Pi~tsch 2
I Institute of Forensic Medicine, Ruprecht-Karls-University, Voss-Str. 2, 69115 Heidelberg, Germany and 2Institute of Forensic
Medicine, Johannes-Gutenberg University, Am Pulverturm 3, 55131 Mainz, Germany
Abstract
Preanalytical stability of a drug and its major metabolites is an
important consideration in pharmacokinetic studies or whenever
the analyte pattern is used to estimate drug habits. Firstly, the
stability of free and glucuronidated 11-nor-Ag-
tetrahydrocannahinol-9-carboxylic acid (THCCOOH,
THCCOOglu) in authentic urine samples was investigated. Random
urine samples of cannabis users (n = 38) were stored at -20, 4, and
20~ up to 15 days and up to 5 days at 40~ and alterations of the
analyte pattern during storage were followed by liquid
chromatography-tandem mass spectrometry. Secondly, the
influence of pH (range 5.0-8.0) on the stability of the analytes was
studied using spiked urine to elucidate the results obtained from
authentic samples. In authentic urine samples, the initial pH ranged
from 5.1 to 8.8. The glucuronide was found to be highly labile at a
storage temperature of 4~ and above. Initially, 18 urine samples
tested positive for THCCOOH. After 2 days storage at 20~
THCCOOH was detectable in a further 4 samples, and 7 more
samples tested positive for THCCOOH (5-81 ng/mL) after 15 days.
Depending on time and temperature, the glucuronide
concentration decreased, resulting in an increase of THCCOOH
concentration. However, a loss in mean total THCCOOH
concentration was found, which was significantly higher in
deteriorated samples than in samples without signs of deterioration
after 15 days of storage at 20~ In the drug-free urine sample
separately spiked with THCCOOglu or THCCOOH, the
investigations on the stability of the target analytes at various
pH values revealed that THCCOOH was stable at pH 5.0. At
higher pH values, its concentration slightly decreased with time,
and about 69% of the initial THCCOOH concentration was
still present at pH 8.0 on day 5. THCCOOglu concentrations
rapidly decreased with increasing pH value. For example,
only 72% of the initial THCCOOglu concentration could
be detected at pH 5.0 on day 1. Degradation of the glucuronide
resulted in formation of THCCOOH, which was observed
even at pH 5.0. In light of the present findings, advanced
9 Author to whom correspondence should be addressed: Priv.-Doz. Dr. Gisela Skopp, Institute
of Forensic Medicine, Voss-Str. 2,69115 Heidelberg, Germany. E-mail:
gisela_skopp@med.uni-heidelberg.de.
forensic interpretations based on the presence of THCCOOH or
the pattern of THCCOOH and THCCOOglu in stored urine
samples seems questionable.
Introduction
Tetrahydrocannabinol is intensively metabolized in man. In
urine, 11-nor-A9-tetrahydrocannabinol-9-carboxylic acid (THC-
COOH) is its major metabolite and is present primarily or ex-
clusively as a glucuronide conjugate (THCCOOglu) (1-5). In
routine casework, cannabis use is usually confirmed by detec-
tion of unconjugated THCCOOH by gas chromatography-mass
spectrometry (GC-MS), which is liberated from its glucuronide
by chemical or enzymatic hydrolysis prior to extraction of a
urine sample.
However, for investigations on pharmacokinetic parameters,
such as renal clearance of conjugated and unconjugated com-
pounds, elimination halflives, and detection windows in urine,
the direct determination of THCCOOH and its glucuronide
should be favored. Especially, when the purpose of the assay is
to use analyte patterns to predict the frequency of cannabis use
(2,5), such an analytical strategy is required. Whenever these
problems are addressed, detailed knowledge on the stability of
the analytes is necessary, for THCCOOglu is an ester-linked [3-
glucuronide known to be a highly labile compound (6).
Many investigations have focused on the stability of THC-
COOH added to drug-free urine specimens (7-12). Only few
data on the analyte pattern and the stability of major cannabi-
noids in authentic urine samples have been reported
(2,5,8,13,14). In a controlled administration study, Alburges
et al. (14) observed that in an infrequent user, all the THCCOOH
excreted in the first 8 h was in the conjugated form, whereas
unconjugated THCCOOH could be detected in urine from a fre-
quent user for at least 24 h. Kelly and Jones (5) reported on free
THCCOOH in urine samples from both infrequent and fre-
quent users, but with lower concentration and shorter detec-
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission. 35

tion time after infrequent drug use. In urine specimens col-
lected from 49 persons up to 10 days while abstaining from
cannabis use, THCCOOH was detectable up to 5 days in infre-
quent users and up to 10 days in frequent users. The glu-
curonide was present in all urine samples up to the end of the
observation period, with higher concentrations in frequent
users. In literature, the pattern of THCCOOH and THCCOOglu
in urine has been suggested to predict recent drug use and to
estimate drug habits, provided a urine sample is available within
1-2 days after the last consumption (2).
The stability of THCCOOglu has been considered in a few
publications, but analysis was restricted to the determination of
free and total THCCOOH (5). Recently, a study on the degra-
dation of THCCOOglu in spiked urine at a constant pH of 6.2
has been published (15). However, it should be remembered
that results obtained from authentic samples are far more com-
plex than may have been expected from experimental data in
the literature (8,13,16,17), for a broad range ofpH values is pre-
sent in urine specimens (18) and the stability of THCCOOglu
has not been thoroughly studied. Therefore, the following in-
vestigations were performed to study time- and temperature-de-
pendent alterations in the pattern of THCCOOH and
THCCOOglu in authentic urine samples and to examine the in-
fluence of pH on the breakdown of THCCOOglu and THC-
COOH under controlled conditions.
Materials and Methods
Chemicals
Deuterated and nondeuterated THCCOOglu (10 IJg/mL,
methanol) was purchased from Alltech Associates (Deerfield,
IL), and deuterated (100 pg/mL, methanol) and nondeuterated
(1 mg/mL, methanol) THCCOOH was supplied by Promochem
(Wesel, Germany). Acetonitrile, methanol, acetic acid, ammo-
nium hydroxide, and ammonium acetate were from Roth (Karl-
sruhe, Germany), and diisopropyl ether was from Merck
(Darmstadt, Germany).
Experimental design
Investigations on the stability of free and glucuronidated
THCCOOH in authentic urine samples. Fresh urine samples
(n = 38, no preservatives) were obtained from volunteers with
a previous history of cannabis use. Subjects were classified as
heavy (n = 24, ___ 1 joint/day) or light users (n = 14, < 1
joint/day) by a questionnaire. The time period between last use
and sampling was less than 48 h. The protocol was approved by
the Ethical Committees of the University Hospital (Heidelberg,
Germany). Aliquots (3 mL) were placed in Teflon-capped vials
(type I glass A) following pH measurement (pH meter Por-
tamess, Knick, Berlin, Germany) and were stored protected
from light up to 15 days at -20, 4, and 20~ and up to 5 days at
40~ The pH was remeasured prior to extraction, and each
specimen was examined for visible signs of deterioration such
as changes in color, transparency, fungal, or bacterial growth.
Experimental investigation on the pH-dependent degrada-
tion of THCCOOglu and THCCOOH. A lyophilized preparation
36
Journal of Analytical Toxicology, Vol. 28, January/February 2004
of human urine (Sigma, Mtinchen, Germany) was dissolved in
distilled water and aliquoted (n = 5, $1-$5). The pH of $1 was
5.6, and pH values of $2-$5 were adjusted to 5.0, 6.0, 7.0, and
8.0 (pH meter, Portamess, Knick), respectively. Specimens were
portioned and fortified either with THCCOOH or its glu-
curonide (--- 400 ng/mL). Twenty aliquots (3 mL) of each of the
spiked samples were placed in Teflon-capped vials (type I glass
A) and stored protected from light up to 5 days at 40~ It had
been shown that THCCOOglu concentration declined according
to an apparent first order reaction and that the kinetic order of
the degradation reaction was the same at 4, 20, and 40~ (15).
In the present experiment, all reaction conditions other than
pH were held constant. Therefore, this accelerated storage
testing was sufficient to investigate the influence of the pH on
the degradation reaction. The initial analyte concentrations
were determined from specimens (n = 4) stored at-20~ within
1 h. Previous studies (15) had assured that THCCOOglu is
stable during storage at -20~ and sample preparation.
Sample preparation and analytical procedure
Extraction was performed as previously described (15).
Briefly, THCCOOglu-d3 and THCCOOH-d 3 (100 ng each) and 25
I~L of acetic acid were added to 500 IJL of urine (samples, stan-
dards). After vortex mixing (60 s), 1000 ILL of diisopropyl ether
was added. The mixture was shaken (20 rain) and centrifuged
(10 rain, 3000 x g). Nine-hundred microliters of the super-
natant was transferred to a silanized vial and evaporated under
nitrogen (40~ The residue was reconstituted in 50 IlL of the
mobile phase.
Analysis was performed using an AP1365 mass spectrometer
with a TurboIon interface operated in the positive ion mode (Ap-
plied Biosystems, Toronto, ON, Canada). The device was inter-
faced to a quarternary pump equipped with an autosampler
(series 200, Perkin Elmer, Oberlingen, Germany). Ten micro-
liters of the processed sample was eluted from a Zorbax Eclipse
XDB C8 column (2.1 150 mm, 5-~m particle size, Hewlett
Packard, Waldbronn, Germany) by acetonitrile/methanol/20mM
ammonium acetate buffer pH 4.0 (41:41:18, v/v) as the mobile
phase at a flowrate of 270 ~L/min.
Quantitation was based on the most prominent precursor to
product ion transitions: for THCCOOglu (-d3), m/z 521 (524)
345 (348) and for THCCOOH (-d3), m/z 345 (348) -~ 327 (330).
The peak-area ratios for the ion transitions of the particular an-
alyte and its internal standard were determined using Mac-
Quan (Masschrom 1.1.1 software, Applied Biosysterns, Toronto,
ON, Canada) and were referenced to the calibration curve. The
six-point calibration curve covered a concentration range of
10-500 ng/mL for each analyte (calibrators: 10, 20, 50, 100, 200,
and 500 ng/mL and a blank). In case of analyte concentration
exceeding the upper level of the calibration curve, the sample
was diluted. Each sample was extracted twice, and each ex-
tract was measured twice. Values given are the means (n = 4).
Degradation of the glucuronide during the analytical proce-
dure could not be observed. Evaluation data (linearity of re-
sponse, limits of detection and quantitation, recovery, and inter-
and intra-assay precision) have been reported (15). The standard
deviation (SD) derived from the between-run precision mea-
surements was 4.1% and 3.7% for THCCOOH and THCCOOglu,

Journal of Analytical Toxicology, Vol. 28, January/February 2004
respectively. SD was used to estimate the stability limits of
THCCOOH and THCCOOglu. The compounds were stable as
long as the difference between their initial concentration (Co)
and their concentration (ct) at a given time (t) did not exceed
the critical difference, d = co- ct < SD (19).
Results and Discussion
Stability of THCCOOglu and THCCOOH in
authentic urine samples
Urine samples stored at -20~ remained unchanged in ap-
pearance, whereas a precipitate was formed at higher storage
temperatures. At a storage temperature of 20~ signs of fungal
or bacterial deterioration were present in about 50% or 79% of
the authentic samples after 10 or 15 days, respectively.
In authentic specimens, the initial pH ranged from 5.1 to 8.8.
In urine stored at -20 or 4~ for 15 days, individual changes
were less than 0.2 pH units. At a storage temperature of 20~
the pI-I steadily increased up to 0.5 pI-I unit in samples that ob-
viously were not deteriorated and differed by as much as 1.5 pH
unit in deteriorated samples after 15 days. Following the pH
susceptibility of acyl glucuronides for hydrolysis (6), such an in-
crease in pH during storage must be suggested to additionally
influence the analyte pattern.
THCCOOglu could be determined in all original specimens,
and THCCOOH was detectable in 18 out of the 38 samples in-
vestigated on day 0. THCCOOH and THCCOOglu were stable up
to 15 days stored at -20~ At other storage temperatures, the
concentration of THCCOOglu steadily decreased with time. As
expected, THCCOOH was formed from THCCOOglu during
storage in a time- and temperature-dependent manner (Table I).
At 20~ THCCOOglu appeared to be stable for one day only.
The distribution of changes in authentic specimens after 2
and 5 days of storage at 4, 20, and 40~ are summarized in Table
II. In samples stored for 2 days at 4~ the concentration of
Table I. Changes in THCCOOgIu and THCCOOH
Concentrations (Range, ng/mL) in Authentic Urine
Samples (n -- 38) During Storage at 4, 20, and 40~
Time of storage (day) Initial 2 5 10 15
4oc
THCCOOglu 8-2780 7-2576 5-2477 5-2417 0-2267
THCCOOH 5-127 0-103 0-170 0-191 0-213
Time of storage (day) Initial 2 5 10 15
20~
THCCOOglu 8-2780 6-2212 9-2213 7-1377 14-903
THCCOOH 5-127 5-184 5-230 6-421 8-1369
Time of storage (day) Initial 1 2 5
40~
THCCOOgIu 8-2780 0-I912 0-1802 0-1209
THCCOOH 5-127 5-487 7-862 7-1252
THCCOOglu remained constant in 6 out of 38 samples. Initially,
18 samples tested positive for THCCOOH, and 10 remained
unchanged in concentration during 2 days of storage at 4~ In
most samples stored at 4 or 20~ for 2 or 5 days, the decrease
of THCCOOglu was 25% or less with reference to the initial
concentration, whereas the amount of unconjugated THC-
COOH increased by up to 99%. On day 2, THCCOOH was pre-
sent in 22 samples stored at 20~ and on day 5, the acid could
be detected in 3 more urine samples. The concentration of
THCCOOH in 11 initially negative tested samples ranged from
5-81 ng/mL on day 15. At a storage temperature of 40~ degra-
dation of THCCOOglu proceeded more rapidly, and in most
samples the initial concentration decreased by a minimum of
-26% within 2 days. Simultaneously, a 2- to 10-fold increase in
THCCOOH concentration could be observed within this time
period. Overall, the trend in the decay of THCCOOglu in au-
thentic samples was similar to that determined from spiked
urine (pH 6.2), where, for example, the glucuronide concen-
tration decreased by 2.6%, 11%, and 26% after 2 days of storage
at 4, 20, and 40~ (15). However, the data from this former in-
vestigation covered only particular points and were insuffi-
cient to explain the broad range of changes observed in
authentic samples.
Table II. Distribution of Changes (% of the Initial
Concentration) in Authentic Urine Samples (n - 38)
After 2 and 5 Days of Storage at 4, 20, and 40~
-1 to -26 to -51 to -76 to
Changes (%) + 0 -25 -50 -75 -100
THCCOOglu--Affer2 Days of Storage
4~ 6 29 3 0 0
20~ 1 26 9 2 0
40~ 0 10 13 8 7
1 to 100 to 500 to
Changes (%) + 0 99 499 999 > 1000
THCCOOH--After 2 Days of Storage
4~ 10 8 0 0 0
20~ 5 9 4 4 0
40~ 0 5 7 7 1
-1 to -26 to -51 to -76 to
Changes (%) +0 -25 -50 -75 -100
THCCOOglu--After 5 Days of Storage
4~ 3 30 4 1 0
20~ 1 18 12 5 2
40~ 0 4 5 11 18
1 to 100 to 500 to
Changes (%) +0 99 499 999 > 1000
THCCOOH--After 5 Days of Storage
4~ 6 11 1 0 0
20~ 3 11 8 3 0
40~ 0 4 7 7 2
37

Data on the stability of THCCOOH from authentic samples
have been variable. Dugan et al. (8) could not establish a trend
in changes of THCCOOH concentration by reanalyzing physio-
logical specimens stored at -20~ whereas Romberg et al. (13)
found a decrease in total THCCOOH concentration, as well as
considerable variability between initial and retested values es-
tablished by GC-MS. Variable recovery following hydrolysis and
extraction may be considered a source of the imprecision. Hy-
drolysis conditions have been optimized to ensure maximal re-
covery of THCCOOH from its glucuronide (20). However,
neither the absolute recovery of liberated THCCOOH has been
established, nor has an analytical procedure such as liquid
chromatography (LC)-MS been used to directly determine
THCCOOH and THCCOOglu. The dynamic changes in the
breakdown of the glucuronide, which have not been considered
in any stability study to date, are suggested to be of considerable
importance for the broad and highly variable changes observed
during storage of auhentic samples.
The results of the present investigation on authentic samples
stored at 20~ were further analyzed by converting the glu-
curonide concentration into molar THCCOOH equivalents and
calculating the total THCCOOH concentration. The sum of the
initial values was used as the reference (100%). In Figure 1, the
mean total (free and conjugated) THCCOOH concentration in
deteriorated samples was compared to that in obviously non-
deteriorated urine specimens. At the end of the observation
period, a significant difference between deteriorated specimens
and samples still in good condition could be established (t-test,
p < 0.05). In addition, an overall loss in total THCCOOH con-
centration was evident, which can be explained by degradation
of THCCOOH rather than by absorption effects, for storage
conditions other than temperature were identical for all sam-
ples. A loss in cannabinoid concentration during storage has
often been attributed to adsorption to material surfaces
(10,21,22). Because the loss of THCCOOH has been shown to
38
0
120
9 deteriorated
100 I-I non deteriorated
~ so
-~ 6O
4o
~ 2o
0
0 2 $ 10 15
Time (day)
Figure 1. Coomparison of the mean total (free and conjugated) THC-
COOH concentration (%) with reference to the mean initial value (100%)
in deteriorated and nondeteriorated authentic samples stored at 20~ for
up to 15 days. * p < 0.05, t-test.
Journal of Analytical Toxicology, Vol. 28, Januaq//FebruaD, 2004
depend on the area of the exposed surface (10), all vials had been
filled with the same volume. Glass was chosen as a container
material in the present investigation. The polar material was
considered suitable to minimize adsorption effects for THC-
COOH and its glucuronide, as both are lipophilic compounds
(23).
Interestingly, Robertson et al. (24) reported that fungal and
bacterial organisms are capable of partially degrading the n-
pentyl sidechain of cannabinoids to yield carboxylic acid and hy-
droxy sidechain derivatives. Therefore, the observed loss of
cannabinoids in deteriorated specimens may partly be attributed
to bacterial or fungal conversion of analytes. In addition, the re-
suits might be explained by the pH, for the ionization of THC-
COOH is sensitive to changes in pH and increases with
increasing pH-favoring decarboxylation of the molecule.
Influence of pH on the stability of THCCOOglu and
THCCOOH in spiked urine ($1-$5)
To support the view that pH strongly influences the analyte
pattern in urine, an experimental study was performed. The re-
suits on the stability of free and glucuronidated THCCOOH in
fortified specimens at different pH values are summarized in
Table III. Initial values were 390 THCCOOH/mL urine and 406
ng THCCOOglu/mL urine. In all series, the particular pH did
not change during the observation period. In the samples spiked
with THCCOOglu, the amount of THCCOOH liberated from
its glucuronide steadily increased at pH 5.0 and 5.6 and was 236
and 218 ng/mL on day 5. At pH 6.0, 7.0, and 8.0, peak concen-
trations of THCCOOH liberated from THCCOOglu were ob-
served on day 3 and were 98, 106, and 130 ng/mL, respectively.
Thereafter, these THCCOOH concentrations decreased by 13%,
25%, and 33%, as compared to peak concentration (100%)
Table III. pH-Dependent Decrease of THCCOOglu and
THCCOOH in Spiked Urine*
Time of storage (day) 1 2 3 4 5
THCCOOglu
pH 5.0 72.3' 36.5* 24.6 t 14.7* detectable t
pH 5.6 46.4* 22.5* 14.0 ~ 10.2' detectable*
pH 6.0 39.4 t 4.3* 0 0 0
pH 7.0 8.3 t detectable t 0 0 0
pH 8.0 9.7* 4.3* 0 0 0
Time of slorage (day) 1 2 3 4 5
THCCOOH
pH 5.0 95.9 98.2 96.6 93.4 89.2
pH 5.6 97.2 96.0 89.1 87.7* 86.2*
pH 6.0 95.7 98.1 84.9 t 78.8* 64.9 t
pH 7.0 97.6 93.6 81.1 ~ 75.8* 65.7 t
pH 8.0 99.6 96.6 85.7* 78.2* 69.2*
* Degradation of THCCOOglu and THCCOOH (expressed as a percentage of the
initial values on day O, 100%1 in urine separately spiked with THCCOOglu
(390 ng/mL! or THCCOOH (406 ng/mL) at pH 5.0, 5.6, 6.0, 7.0, and 8.0 at
40~ means in = 4).
t > critical difference, d = Co - q < SD, where ca is the initial concentration, c t is
the concentration at a given lime, t is the time, and SD is the standard deviation
derived from between-run precision measurement (19).

Journal of Analytical Toxicology, Vol. 28, January/February 2004
during the following 48 h. In samples that were separately
spiked with THCCOOH, the analyte concentration was found to
be most stable at pH 5.0, whereas it continuously decreased at
pH 5.6 or above.
Degradation of THCCOOglu in body fluids has been sug-
gested to occur by enzymatic or chemical hydrolysis (25). The
present results indicated that spontaneous hydrolysis of the
glucuronide in urine is strongly dependent on pH. In 1983,
Law et al. (1) reported that hydrolysis of THCCOOglu occurred
more rapidly in an alkaline urine sample stored at room tem-
perature. Therefore, the authors speculated that free THC-
COOH in urine might be an artifact rather than a physiological
finding. Meanwhile, it has been substantiated by LC-MS-MS
analysis that unconjugated THCCOOH is present in authentic
urine samples in minute amounts up to 7% a few days after ab-
staining from cannabis use (2). However, the present results
showed that THCCOOH and THCCOOglu, as well as total THC-
COOH concentrations, might undergo dynamic changes in
urine samples depending on pH and storage conditions.
Conclusions
The present investigation demonstrated the need to study
the stability of the target analyte including its glucuronide
metabolite in authentic samples to cover the spectrum of al-
terations that might occur during transport and short-term
storage. THCCOOH was not detectable in 20 out of 38 fresh, au-
thentic urine samples. Positive findings were observed in 18
specimens and constituted up to 7% of the total amount of
THCCOOH. At the end of the observation period, 5--81 ng THC-
COOH/mL urine was detectable in 11 samples initially tested
negative. During storage, THCCOOH was liberated from its
glucuronide in a time- and temperature-dependent manner,
and individual changes showed a broad range. Stability of THC-
COOglu was far more influenced by pH than THCCOOH.
Overall, a loss in total THCCOOH concentration could be ob-
served, which may be attributed to decarboxylation of THC-
COOH and partially to fungal or bacterial conversion of the
analytes. In deteriorated specimens stored for 15 days at 20~
the loss in total THCCOOH concentration was significantly
higher compared to samples that were in good condition.
From the results obtained from spiked urine, it can be con-
cluded that acidification to pH 5 may help to enhance the sta-
bility of the analytes, but liberation of THCCOOH from its
glucuronide cannot be completely prevented. These results also
confirmed the influence of time and temperature on the analyte
pattern, but underlined that experimental data should not be
transferred to forensic material without restrictions. In the
present investigation, a complete loss of the analytes did not
occur (Tables I and II), therefore confirmation of cannabis use
seems promising, even in unfavorably stored samples (e.g., at
20~ for 15 days). However, information on the history of the
sample may be required for pharmacokinetic or advanced
forensic interpretation. Although a correlation of THCCOOH
and its glucuronide to estimate the time of cannabis use must
fail due to frequently negative THCCOOH findings or very low
THCCOOH concentrations (Table I), previous investigations on
THCCOOglu indicated that a differentiation between infrequent
and frequent use of cannabis, often requested by the German
authorities, might be promising (2). Also, the detection of un-
conjugated THCCOOH in urine has been suggested as an indi-
cation of heavy cannabis use (14). In light of the present
findings, its reliability as a marker for frequent drug consump-
tion seems questionable, for usually the transport of a urine
specimen from the collection side to the forensic laboratory re-
quires two days, at least.
Acknowledgments
This work was supported in part by the Deutsche Forschungs-
gemeinschaft (Grant Sk 48/2-2). The authors thank S. Ros-
tock-Wolf for her technical assistance.
References
1. B. Law, RA. Mason, A.C. Moffat, R.I. Gleadle, and L.J. King.
Forensic aspects of the metabolism and excretion of cannabinoids
following oral ingestion of cannabis resin. J. Pharm. Pharmacol. 36:
289-294 (1984).
2. G. Skopp, L. P6tsch, B. Ganssmann, M. Mauden, B. Richter, R.
Aderjan, and R. Mattern. Freie und glucuronidierte cannabinoide
im urin--untersuchungen zur einsch~tzung des konsumverhal-
tens. Rechtsmed 10:21-28 (1999).
3. S. Agurell, M. Halldin, J.E. Lindgren, A. Ohlsson, M. Widman, H.
Gillespie, and L. Hollister. Pharmacokinetics and metabolism of A~-
tetrahydrocannabinol and other cannabinoids with emphasis on
man. PharmacoL Rev. 38:21-43 (1986).
4. RL. Williams and A.C. Moffatt. Identification in human urine ofg 9-
tetrahydrocannabinol-11-oic acid glucuronide: a tetrahydro-
cannabinol metabolite. J. Pharm. Pharmacol. 32" 445-448 (1980).
5. R Kelly and R.T. Jones. Metabolism of tetrahydrocannabinol in fre-
quent and infrequent marijuana users. J. Anal. Toxicol. 16:
228-235 (1992).
6. B.C. Sallustio, L. Sabordo, A.M. Evans, and R.L. Nation. Hepatic
disposition of electrophilic acyl glucuronide conjugates. Curt.
Drug Metab. 1 9 163-180 (2000).
7. A. Smith-Kielland, B. Skuterud, and J. Morland. Urinary excretion
of 11-nor-9-carboxy-A%tetrahydrocannabinol and cannabinoids
in frequent and infrequent drug users. ]. Anal. ToxicoL 23:323-332
(1999).
8. S. Dugan, S. Bogema, R.W. Schwartz, and N.T. Lappas. Stability of
drugs of abuse in urine samples stored at -20~ J. Anal Toxicol.
18:391-396 (1994).
9. G. Anderson and M. Scott. Determination of product shelf life
and activation energy for five drugs of abuse. Clin. Chem. 37:
398-402 (1991 ).
10. K.D.W. Roth, N.A. Siegel, R.W. Johnson, L. Litauszki, L. Salvati,
C.A. Harrington, and L.K. Wray. Investigation of the effects of so-
lution composition and container material type on the loss of 11 -
nor-A%THC-9-carboxylic acid. J. Anal. Toxicol. 20:291-300
(1996).
11. D.E. Moody, K.M. Monti, and A.C. Spanbauer. Long-term stability
of abused drugs and antiabuse chemotherapeutical agents stored
at -20~ J. AnaL ToxicoL 23:535-540 (1999).
12. B.D. Paul, R.M. McKinley, J.K. Walsh, T.S. Jamir, and M.R. Past. Ef-
fect of freezing on the concentration of drugs of abuse in urine.
J. Anal ToxicoL 17:378--380 (1993).
39

13. R.W. Romberg and M.R. Past. Reanalysis of forensic urine speci-
mens containing benzoylecgonine and THC-COOH. J. Forensic
Sci. 39:479-485 (1994).
14. M.E. Alburges and M.A. Peat. Profiles of delta-9-tetrahydro-
cannabinol metabolites in urine of marijuana users: preliminary
observations by high performance liquid chromatography-
radioimmunoassay. J. Forensic Sci. 31:695-706 (1986).
15. G. Skopp and L. P6tsch. Stability of 11-nor-9-carboxy-Ag-tetrahy -
drocannabinol glucuronide in plasma and urine assessed by liquid
chromatography-tandem mass spectrometry. Clin. Chem. 48:
301-303 (2002).
16. B. Levine and M.L. Smith. Stability of drugs of abuse in biological
specimens. Forensic Sci. Rev. 2:147-157 (1990).
17. S. Golding Fraga, J.F. Diaz-Flores Estevez, and C. Diaz Romero.
Stability of cannabinoids in urine in three storage temperatures.
Ann. Clin. Lab. Sci. 28:160-162 (1998).
18. W.G. Guder. Niere und ableitende Harnwege. In Lehrbuch der
Klinischen Chemie und Pathobiochemie, H. Greiling and A.M.
Gressner, Eds. Schattauer, New York, NY, 1995, pp 743-744.
19. D. Stature. A new concept for quality control of clinical laboratory
investigations in the light of clinical requirements and based on ref-
erence method values. J. Olin. Chem. Olin. Biochem. 20:817-824
(1982).
20. P.M. Kemp, I.K. Abukhalaf, J.E. Manno, B.R. Manno, D.D. Alford,
and G.A. Abusada. Cannabinoids in humans. I. Analysis of Ag.
40
Journal of Analytical Toxicology, Vol. 28, January/February 2004
tetrahydrocannabinol and six metabolites in plasma and urine
using GC-MS. J. Anal. Toxicol. 19:285-291 (1995).
21. T. Dextraze, W.C. Griffiths, R Camara, L. Audette, and M. Rosner.
Comparison of fluorescence polarization immunoassay, enzyme
immunoassay and thin-layer chromatography for urine cannabi-
noid screening: effects of analyte adsorption and vigorous mixing
of specimen on detectability. Ann. Clin. Lab. Sci. 19:133-138
(1989).
22. W.A. Joern. Surface adsorption of the urinary marijuana carboxy
metabolite: the problem and a partial solution. J. Anal. ToxicoL 16:
401 (1992).
23. G. Skopp, L. P6tsch, M. Mauden, and 8. Richter. Partition coeffi-
cient, blood to plasma ratio, protein binding and short-term sta-
bility of 11-nor-Ag-carboxy tetrahydrocannabinol glucuronide.
Forensic Sci. Int. 126:17-23 (2002).
24. L.W. Robertson, S.W. Koch, S.R. Huff, R.K. Malhotra, and A. Gosh.
Microbiological oxidation of the pentyl side chain of cannabi-
noids. Experientia 34:1020-1022 (1978).
25. S.W. Toennes and G.F. Kauert. Importance of Vacutainer selection
in forensic toxicological analysis of drugs of abuse. J. Anal. ToxicoL
25:339-343 (2001).
Manuscript received August 12, 2002;
revision received June 19, 2003.