Beck_jatja11.qxd:JATLynneTemplate
9/7/11
9:58 AM
Page 1
Journal of Analytical Toxicology, Vol. 35, October 2011
Detection of ∆9-Tetrahydrocannabinol in Exhaled Breath
Collected from Cannabis Users
Olof Beck1,*, Sören Sandqvist1, Ilse Dubbelboer2, and Johan Franck3
1Department
of Medicine, Section of Clinical Pharmacology, Karolinska Institutet, Stockholm, Sweden; 2Laboratory for TDM
and Clinical Toxicology, Department of Pharmacy, University Medical Center Groningen, Groningen, The Netherlands; and
3Department of Clinical Neuroscience, Division of Psychiatry, Karolinska Institutet, Stockholm, Sweden
Abstract
Exhaled breath has recently been proposed as a new possible
matrix for drugs of abuse testing. A key drug is cannabis, and the
present study was aimed at investigating the possibility of detecting
tetrahydrocannabinol and tetrahydrocannabinol carboxylic acid in
exhaled breath after cannabis smoking. Exhaled breath was
sampled from 10 regular cannabis users and 8 controls by directing
the exhaled breath by suction through an Empore C18 disk. The disk
was extracted with hexane/ethyl acetate, and the resulting extract
was evaporated to dryness and redissolved in 100 µL hexane/ethyl
acetate. A 3-µL aliquot was injected onto the LC–MS–MS system
and analyzed using positive electrospray ionization and selected
reaction monitoring. In samples collected 1–12 h after cannabis
smoking, tetrahydrocannabinol was detected in all 10 subjects. The
rate of excretion was between 9.0 and 77.3 pg/min. Identification
of tetrahydrocannabinol was based on correct retention time
relative to tetrahydrocannabinol-d3 and correct product ion ratio.
In three samples, peaks were observed for tetrahydrocannabinol
carboxylic acid, but these did not fulfill identification criteria.
Neither tetrahydrocannabinol or tetrahydrocannabinol carboxylic
acid was detected in the controls. These results confirm older
reports that tetrahydrocannabinol is present in exhaled breath
following cannabis smoking and extend the detection time from
minutes to hours. The results further support the idea that exhaled
breath is a promising matrix for drugs-of-abuse testing.
Introduction
Exhaled breath has been proposed as a new possible matrix for
drugs of abuse testing (1–4). Following the first demonstration
of amphetamine and methamphetamine being detectable in
breath after drug intake, methadone has been used as an experimental compound to further develop and validate the breath
sampling procedure. The procedure using a C18 filter intended
for solid-phase extraction has been found to be better than collecting exhaled breath condensate and is also easier to perform
in practice (4). This procedure was validated for methadone and
* Author to whom correspondence should be addressed. Email: [email protected].
found to be reproducible from a qualitative perspective. The
possibility to detect drug use by using exhaled breath is intriguing when considering that alcohol testing technology has
been developed to the point that on-site breath testing with
legally defensible results using infrared spectroscopy can be
performed and also used for vehicle alcolocks (5,6).
Four previous papers from over 25 years ago describe the detection of (–)-∆9-tetrahydrocannabinol (THC) in breath after
cannabis smoking (7–10). Thereafter, the idea of detecting
drugs of abuse in exhaled breath seems to have been forgotten
until recently. In the study by Manolis and co-workers (7), different sampling procedures were examined. The final instrumental analysis was performed with capillary column gas chromatography–mass spectrometry (MS) providing detectability
of 250 pg THC. The best trapping method was found to be
Tenax GC tubes with molecular sieves. At the time point 10–12
min post smoking, THC was detectable in 10 out of 14 subjects,
but at 20 min, no breath sample contained detectable THC.
The main metabolite of THC, (±)-11-nor-carboxy-∆ 9 tetrahydrocannabinol carboxylic acid (THCA), was not detected
in any sample.
In order to further explore exhaled breath as a matrix for
drug testing, we undertook the present investigation of
cannabis detection in order to reproduce and expand the earlier reported results. The present study was performed with an
alternative sampling procedure and with instrumentation providing increased sensitivity.
Materials and Methods
Chemicals and materials
THC, THC-d3, THCA, and THCA-d9 were obtained as ampouled methanol solutions from Cerilliant (Round Rock, TX).
Methanol, acetonitrile, and ethyl acetate of HPLC grade were
from JT Baker (Mallinckrodt Baker BV, Deventer, Holland).
Formic acid of HPLC grade was from Merck GmbH (Darmstadt,
Germany). Hexane of HPLC grade was from VWR Interna-
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.
541
Beck_jatja11.qxd:JATLynneTemplate
9/7/11
9:59 AM
Page 2
Journal of Analytical Toxicology, Vol. 35, October 2011
tional (West Chester, PA). The Milli-Q water was of ultra-pure
quality (>18MΩ/cm) and prepared in-house. The analytical
column, AQUITY UPLC BEH C18 (1.7 µm, 1.0 × 100 mm), was
from Waters (Milford, MA). The 47-mm C18 Empore disk was
from Varian (Palo Alto, CA).
Preparation of stock solutions
The ampouled THC, THC-d3, THCA, and THCA-d9 served as
stock solutions. These solutions were further diluted to suitable
concentrations using methanol and finally with 20% of 0.1%
formic acid and stored at –18°C for a maximum of 1 year.
Patients and control subjects
Ten patients who are regular cannabis users were recruited
for sample collection (ages 20–55). Ethical approval was obtained from the Stockholm Regional Ethics Review Board
(No 2008/1347-31). As a control group, eight healthy, drugfree volunteers (5 males, 3 females, aged 26–69) were recruited. Written informed consent was obtained from all participants.
MS analysis
A 3-µL aliquot was subjected to analysis by selected reaction
monitoring (SRM) ultra-performance liquid chromatography
(UPLC)–MS–MS (Waters Quattro Premier XE). The chromatographic system included an AQUITY UPLC BEH C18
column (100 mm × 1.0 mm, particle size 1.7 µm) with a gradient system consisting of A = 0.1% formic acid and B = acetonitrile. The mobile phase was 95% A for 1.2 min, followed by
a linear gradient from 5% B to 65% B to 3.0 min. The equilibration time between injections was 4.0 min (95% A). The
flow rate was 0.20 mL/min.
Two product ions from the protonated molecules were monitored for THC and THCA, and one was monitored for THC-d3
and THCA-d9 (Table I). This was done by SRM in the positive
electrospray mode, with 75 ms dwell time for each channel.
Other instrumental settings are provided in Table I.
Quantification
Standards for quantification were prepared from fortified
blank Empore disks. These were prepared by adding 50–4000
pg of THC and THCA on the surface. After drying the disks were
Sampling of exhaled breath
Compounds present in the exhaled breath were collected
for 10 min by suction through a 47-mm Empore C18 disk using
a membrane pump to assist the flow (pump capacity 300
mL/min). The subjects were asked to breathe more deeply than
normal into a mouthpiece (No. 4091148, Palmenco AB, Stockholm, Sweden) mounted in the sampling device holding the
Empore disk (4). It was estimated that almost all the exhaled
breath was collected through the filter during the sampling period. Following sampling, the Empore disk was dismantled
using a tweezers and stored at –80°C. Careful cleaning of the
sampling device, which takes about 15 min, was done between
samplings using 70% ethanol.
Sample preparation
Table I. Instrumental Settings of the Mass Spectrometer*
Analyte
THC
THC
THC-d3
THCA
THCA
THCA-d9
Precursor Ion Product Ion Cone Voltage
m/z
m/z
(V)
315.4
315.4
318.4
343.4
343.4
352.2
193.1
122.9
196.0
299.3
245.2
308.2
Collision
Energy (eV)
32
32
37
47
47
47
* Source block temperature, 120°C; desolvation gas temperature, 450°C;
desolvation gas flow, 950 L/h; cone gas flow, 50 L/h; capillary voltage,
3.4 kV for THC and 2.0 kV for THCA; collision gas flow, 0.30 mL/min;
and multiplier voltage 700 V for THC and 690 V for THCA.
23
35
23
23
29
23
Following storage, the Empore disk was cut into 5- × 5-mm
pieces using a scalpel, and all pieces were
transferred to a 10 mL glass test-tube. A
volume of 10 or 25 µL of 40 ng/mL THCTable II. Summary of Data Obtained for THC and THCA in Exhaled Breath from
d3 and THCA-d9 was added and mixed
10 Patients
using a vortex mixer, 300 µL of 2-propanol
was added (to wet the surface), mixed, and
Time after
THC
THCA
Subject
Number of
Cannabis Use
Excretion
Excretion
finally 5 mL of hexane/ethyl acetate (4:1)
No.
Breaths
Sex
(h)
(pg/min)*
(pg/min)*
was added. This mixture was shaken for 1
h in a thermostatic bath at 37°C. There1
38
Male
1
67.6
Detected†
after, the test-tube was centrifuged for 15
2
29
Male
1
29.3
< 2.5
min at 3000 × g at 10°C, the supernatant
3
30
Male
1
28.3
< 2.5
transferred to a new 10-mL glass test4
33
Male
1
23.3
< 2.5
tube, and the extraction procedure re5
38
Female
1
23.6
< 2.5
peated using 1 mL of hexane/ethyl acetate
6
24
Male
1
18.0
< 2.5
(4:1). Finally, the two supernatants were
7
59
Male
1
77.3
Detected†
combined, 10 µL of 10% aqueous formic
8
36
Male
2
46.6
< 2.5
acid added, and evaporated to dryness
9
33
Female
2
33.3
< 2.5
10
56
Male
12
9.0
Detected†
under a stream of nitrogen at a temperature of 40°C. The dry residue was dis* Water was used to rinse th.e mouth before sampling. Total sampling time was 10 min.
solved in 100 µL of hexane/ethyl acetate
† Peaks were present at transition 343.4 → 299.3 above LOD.
(1:1).
542
Beck_jatja11.qxd:JATLynneTemplate
9/7/11
9:59 AM
Page 3
Journal of Analytical Toxicology, Vol. 35, October 2011
prepared for analysis as described. Calibration curves were
constructed using linear regression analysis, with weighting
factor 1/x (x = concentration). The transitions used for quantifications were 315.4 → 193.1 for THC and 343.4 → 299.3 for
THCA.
A
Method validation
Five replications of the calibration curve were analysed on
different occasions. Limit of detection (LOD) was assessed by
applying 3.75 pg of THC and 7.5 pg of THCA onto a blank Empore disk and subjecting it to analysis. Recoveries of THC and
THCA from the filter surface were estimated by fortifying the
filter surface with reference solutions (80–1600 pg) and subjecting these to analysis.
B
Results
Method validation
Both compounds gave linear response in the calibration
curves (n = 5) with correlation coefficients ranging between
0.956 and 0.998 (mean 0.986) for THC and 0.888 and 0.999
(mean 0.943) for THCA. The LODs were 2.5 pg on column for
THC and 1.5 pg for THCA. Recoveries of THC and THCA from
filter surface were 96% ± 9 (SD, n = 5) and 100% ± 11 (SD,
n = 5), respectively.
C
Analysis of exhaled breath from cannabis smokers
THC was detected in sampled exhaled breath from 10 studied
patients (Table II). Nine of these subjects were sampled within
a few hours after smoking. In these subjects, the time after
smoking was reported to be 1 or 2 h; THC was detected and
quantified in all of the samples. The amount of THC ranged between 18.0 and 77.3 pg/min. THC was detectable in one patient’s
sample collected about 12 h after smoking. The identity of THC
was supported by correct retention times relative to internal
standard and with correct product ion ratios. In three of these
subjects, a peak above the LOD was observed for THCA, but only
for one transition. Therefore, the detection of THCA did not
meet identification criteria. Chromatograms for a calibration
sample for THC and THCA and a blank is shown in Figures 1A–
1C. Chromatogram from two authentic samples (cases 1 and 7
of Table II) is shown in Figures 1D and 1E. No THC or THCA
was detected in exhaled breath from the eight healthy volunteers used as controls.
D
E
Discussion
The results of this study support the recent suggestion that
exhaled breath can be used for detection of drugs of abuse. It
confirms an earlier report that THC can be detected in exhaled breath following cannabis smoking (7) and extends the
detection window from 12 min to possibly 12 h.
Breath testing is an established field in medical research
with clinical applications for detection of cancer, infection, di-
Figure 1. Chromatogram for a THC standard containing 200 pg/sample (A);
chromatogram for a THCA standard containing 200 pg/sample (B);
chromatogram for a THC system blank sample (C); chromatogram for
THC in a patient sample (case 1 of Table II) containing 676 pg/sample (D);
and chromatogram for THC in a patient sample (case 7 of Table II) containing 773 pg/sample (E). The chromatograms represent raw data with no
smoothing applied.
543
Beck_jatja11.qxd:JATLynneTemplate
9/7/11
9:59 AM
Page 4
Journal of Analytical Toxicology, Vol. 35, October 2011
abetes, liver disorder, and asthma (11). It is estimated that
human breath contains up to 3000 compounds (12). These
comprise volatile compounds carried in the vapor phase and
non-volatile compounds assumed to be carried in aerosol particles. The aerosol phase can be collected as exhaled breath condensate and is known to contain non-volatile metabolites and
even proteins (12,13). It has been demonstrated by specifically
collecting breath aerosol particles that substances typical for
the airway lining fluid (surfactant) are exhaled as aerosol particles (14). It is therefore not surprising that compounds of exogenous origin are also present in exhaled breath, especially
with regard to THC because it is administered by inhalation.
On the other hand, amphetamine, methamphetamine, and
methadone that are also present in breath most likely originate
from the circulation. Saliva contamination is of concern in
breath analysis, and it cannot be excluded that this occurred
also in this study. However, when using a saliva trap system, as
was the case in our device, it has been shown that saliva contamination is trivial (15).
In recent years, drugs of abuse testing using oral fluid has
been developed for forensic and clinical applications (16). There
is an interest for finding alternative matrices to urine and
blood. Breath testing can potentially offer an attractive alternative as it is easy and safe to collect and is a non-invasive sampling procedure. The sampling of exhaled breath might also be
safe from an adulteration point of view.
The sampling procedure used in the present work was based
on the collection of compounds on a modified silica surface.
Because of the back-pressure, the flow of air over the filter was
supported by pumping. The Empore C18 disk is a filter intended for use when extracting compounds from larger volumes of water. However, this filter has also found use in trapping compounds (insecticides) from indoor air (17). It has
been shown that the filter is effective in trapping pyrethroids
from indoor air (17). For methadone, we have found the filter
procedure being more effective than collecting exhaled breath
condensate (2,3).
Of the four previous studies (7–10), only one (7) reported
detection of THC in breath using a reliable detection method
(i.e., MS). Our results are in full agreement with the report of
Manolis and co-workers (7) in the detection THC but not
THCA. We found a longer detection time, but this can simply
be explained by the lower detection level of our method.
Detection of cannabis use can be considered a key parameter
in testing for drugs of abuse. The demonstration of this study
that THC is detectable for several hours post cannabis smoking
further supports the possibility that exhaled breath is suitable
for drugs of abuse testing.
Acknowledgments
We thank Inger Engman-Borg for assistance in the clinical
part if this work. This work was supported by grants from The
Swedish Governmental Agency for Innovation Systems—
544
Vinnova, the Swedish Research Council, and the Stockholm
County Council.
References
1. O. Beck, K. Leine, G. Palmskog, and J. Franck. Amphetamines detected in exhaled breath from drug addicts: a new possible
method for drugs-of-abuse testing. J. Anal. Toxicol. 34: 233–237
(2010).
2. O. Beck, S. Sandqvist, P. Eriksen, J. Franck, and G. Palmskog.
Method for determination of methadone in exhaled breath collected from subjects undergoing methadone maintenance treatment. J. Chromatogr. B 878: 2255–2259 (2010).
3. O. Beck, S. Sandqvist, P. Eriksen, J. Franck, and G. Palmskog. Determination of methadone in exhaled breath condensate by liquid
chromatography–tandem mass spectrometry. J. Anal. Toxicol. 35:
129–133 (2011).
4. O. Beck, S. Sandqvist, M. Böttcher, P. Eriksen, J. Franck, and
G. Palmskog. Study on the sampling of methadone from exhaled
breath. J. Anal. Toxicol. 35: 257–263 (2011).
5. L. Lindberg, S. Brauer, P. Wollmer, L. Goldberg, A.W. Jones, and
S.G. Olsson. Breath alcohol concentration determined with a
new analyzer using free exhalation predicts almost precisely the
arterial blood alcohol concentration. Forensic Sci. Int. 168: 200–
207 (2007).
6. B. Hök, H. Pettersson, A. Kaisdotter-Andersson, S. Haasl, and
P. Åkerlund. Breath analyzer for alkolocks and screening devices.
IEEE Sensors J. 10: 10–15 (2010).
7. A. Manolis, L.J. McBurney, and B.A. Bobbie. The detection of ∆9tetrahydrocannabinol in the breath of human subjects. Clin
Biochem. 16: 229–233 (1983).
8. T.J. McCarthy and J.D. van Zyl. Breath analysis of cannabis
smokers. J. Pharm. Pharmacol. 24: 489 (1972).
9. G. Hauck and H.R. Moll. Versuche zum nachweis von Cannabisinhaltstoff in der ausatenluft. Beitrage Gericht. Medizin 23: 221–
226 (1974).
10. J.L. Valentine, P.J. Bryant, P.L. Gutshall, H.M.G. Owen, and
H.C. Niu. Detection of ∆9-tetrahydrocannabinol in human breath
following marijuana smoking. Anal. Lett. 12: 867–880 (1979).
11. A. Mashir and R.A. Dweik. Exhaled breath analysis: the new interface between medicine and engineering. Adv. Powder Technol.
20: 420–425 (2009).
12. B.B. Buszewski, M. Kesy, T. Ligor, and A. Amann. Human exhaled
air analytics: biomarkers of disease. Biomed. Chromatogr. 21:
553–566 (2007).
13. W. Miekisch and J.K. Schubert. From highly sophisticated analytical techniques to life-saving diagnostics: technical developments in breath analysis. Trends Anal. Chem. 25: 665–673 (2005).
14. A.C. Almstrand, E. Ljungström, J. Lausmaa, B. Bake, P. Sjövall, and
A.C. Olin. Airway monitoring by collection and mass spectrometric analysis of exhaled particles. Anal Chem. 81: 662–668
(2009).
15. P. Montuschi. Analysis of exhaled breath condensate in respiratory
medicine: methodological aspects and potential clinical applications. Ther. Adv. Respir. Dis. 1: 5–23 (2007).
16. W.M. Bosker and M. Huestis. Oral fluid testing for drugs of abuse.
Clin. Chem. 55: 1910–1931 (2009).
17. T. Yoshida. Simultaneous determination of 18 pyrethroids in indoor air by gas chromatography/mass spectrometry. J. Chromatogr.
A 1216: 5069–5076 (2009).
Manuscript received November 12, 2010;
revision received January 11, 2011.