Drug Testing

Research article and Analysis

Received: 18 February 2014 Revised: 8 May 2014 Accepted: 14 June 2014 Published online in Wiley Online Library: 28 July 2014

(www.drugtestinganalysis.com) DOI 10.1002/dta.1692

Cannabinoid findings in children hair what do

they really tell us? An assessment in the light of
three different analytical methods with focus
on interpretation of 9-tetrahydrocannabinolic
acid A concentrations
Bjoern Moosmann,a,b Nadine Roth,a,b Martin Hastedt,c
Andrea Jacobsen-Bauer,d Fritz Pragstc and Volker Auwrtera*
Hair analysis for drugs and drugs of abuse is increasingly applied in child protection cases. To determine the potential risk to a
child living in a household where drugs are consumed, not only can the hair of the parents be analyzed but also the hair of the
child. In the case of hair analysis for cannabinoids, the differentiation between external contamination and systemic uptake is
particularly difcult, since the drug is quite often handled extensively prior to consumption (e.g. when preparing a joint) and
smoke causes a further risk for an external contamination. 9-tetrahydrocannabinolic acid A (THCA-A), the non-psychoactive
biogenetic precursor of 9-tetrahydrocannabinol (THC), is a suitable marker for external contamination since it is not
incorporated into the hair matrix through the bloodstream in relevant amounts. In the presented study, hair samples from
41 children, 4 teenagers, and 34 drug-consuming parents were analyzed for THCA-A, THC and cannabinol (CBN) applying meth-
anolic extraction and a fully validated liquid chromatography-tandem mass spectrometry (LC-MS/MS) method (Method 1). For
comparison, a part of the samples was also analyzed applying alkaline hydrolysis followed by liquid/liquid extraction and gas
chromatography-mass spectrometry (GC-M)S (Method 2), or by headspace-solid phase microextraction-gas chromatography-
mass spectrometry (HS-SPME-GC-MS) (Method 3). Furthermore, 458 seized marihuana samples and 180 seized hashish samples
were analyzed for the same cannabinoids by gas-chromatography-ame ionization detector (GC-FID). In all but one of the hair
samples, the concentration of THCA-A was higher than the concentration of THC and in 14 cases no THC could be detected
despite the presence of THCA-A, suggesting that in almost all cases a signicant external contamination had occurred.
Within-family comparison showed a higher THCA-A/THC ratio in hair of children than of their consuming caregivers. Mean
and median of this ratio of all hair samples (6.7 and 4.2) were between those of marihuana (11.0 and 8.3) and hashish (2.8
and 2.1) with a large variation in all samples. Comparison of the Methods 1 to 3 showed clearly that the choice of the analytical
procedure has a strong inuence on the quantitative results, mainly because of decarboxylation of THCA-A during hair
hydrolysis by NaOH and other analytical steps, which lead to artifactually elevated THC concentrations. In conclusion, these
ndings suggest that the major part of the cannabinoids detected in the hair samples from children arose from an external
contamination through passive transfer by e.g. contaminated hands or surfaces and not from inhalation or deposition of side
stream smoke. Copyright 2014 John Wiley & Sons, Ltd.
Additional supporting information may be found in the online version of this article at the publishers web site.

Keywords: child protection; external contamination; hair analysis; tetrahydrocannabinol; tetrahydrocannabinolic acid A

Introduction
Analysis of hair samples for drugs of abuse is applied in various
settings today and has become a standard method for certain
issues,[1] three of the most common being child protection
cases,[24] abstinence control programs,[57] and workplace drug
testing.[8] Considering the severe consequences positive or nega-
tive test results may have for the person concerned, inter-
pretation has to be carried out with utmost care. In particular,
external contamination should be excluded or discussed ade-
quately. In the case of hair analysis for cannabinoids this is partic-
ularly difcult, since the drug is quite often handled extensively
prior to consumption (e.g. when preparing a joint) and smoke
* Correspondence to: V. Auwrter, Institute of Forensic Medicine, Forensic
Toxicology Department, University Medical Center Freiburg, Albertstrae 9,
79104 Freiburg, Germany. E-mail: volker.auwaerter@uniklinik-freiburg.de
a Institute of Forensic Medicine, Forensic Toxicology Department, University
Medical Center Freiburg, Albertstrae 9, 79104, Freiburg, Germany
b Hermann Staudinger Graduate School, University of Freiburg, Hebelstrae 27,
79104, Freiburg, Germany
c Institute of Legal Medicine, Charit-University Medicine Berlin, Turmstrae 21,
Building N, 10559, Berlin, Germany
d State Ofce of Criminal Investigation Baden-Wrttemberg, Taubenheimerstrae
349

causes a further source of external contamination. Further 85, 70372, Stuttgart, Germany

Drug Test. Analysis 2015, 7, 349357 Copyright 2014 John Wiley & Sons, Ltd.
 Drug Testing
and Analysis
aggravating the interpretation of THC positive hair results is
the circumstance that quite often the oxidative metabolite
9-carboxy-11-nor-9-tetrahydrocannabinol (THC-COOH), which
would conrm a body passage, cannot be detected despite
considerable 9-tetrahydrocannabinol (THC) concentrations
and application of highly sensitive analytical methods.[916]
A previous study has shown that 9-tetrahydrocannabinolic
acid A (THCA-A, structure in Figure 1), the biogenetic non psycho-
active precursor of THC, could potentially serve as a marker for an
external hair contamination, since it is not incorporated into hair
through the bloodstream after repeated oral intake of high doses
THCA-A.[17] However, in many forensic hair samples analyzed in
our laboratory, THCA-A was detected in concentrations ex-
ceeding that of THC (own unpublished data and Auwrter
et al.[17]). These observations lead to the assumption that THCA-
A in hair results from external contamination exclusively. Addi-
tionally, also parts of the detected THC and cannabinol (CBN)
may originate from such sources. Based on further experiments
it could be ruled out with a high certainty that relevant amounts
of THCA-A are transferred through side-stream marihuana smoke,
as the smoke only contains negligible amounts of THCA-A.[18]
One possible way of an external contamination of hair could
also be transfer through contaminated hands[19] as it has already
been shown in a contamination experiment for heroin.[20] Given
the high lipophilicity of cannabinoids, it seems plausible that
during the handling of raw cannabis plant material or hashish -
as it occurs when e.g. rolling a joint or preparing hash-cookies
a transfer to the ngers and later to the own or the childrens hair
might occur. Such a contamination may take place for instance
when cannabis consuming parents, relatives or acquaintances
sweep their hands through the hair of a young child.
Hair analysis for cannabinoids is further complicated by
the fact that quite often alkaline hydrolysis is used as the
method of sample preparation,[17,2124] leading to decarbox-
ylation of THCA-A and therefore articially elevating the THC
concentration.
Regarding child protection cases, the latest report from the
European Monitoring Centre for Drugs and Drug Addiction
(EMCDDA) states that about 10% of clients entering treatment
for drug problems in 2010 lived with children.[25] It is often help-
ful to evaluate whether the children are directly exposed to drugs
in order to evaluate health risks. In cases of positive ndings in
children hair it has to be carefully differentiated whether (1) the
drug was administered to the child for example for sedating it,
(2) the drug was consumed in the presence of the child leading
to passive uptake by the child, or (3) the drug was consumed in
the absence of the child. All three scenarios bear different health
risks for the child and probably require different approaches from
a child-care perspective.
In order to evaluate based on the presence of THCA-A, if the
positive cannabinoid results found in children hair in a child
protection project[3] are caused by external contamination, hair sam-
ples from 41 children and 34 cannabis-consuming parents were
9
Figure 1. Formation of -tetrahydrocannabinol (THC) by decarboxylation
350
9
of -tetrahydrocannabinolic acid A (THCA-A).
wileyonlinelibrary.com/journal/dta Copyright 2014 John Wiley & Sons, Ltd.
B. Moosmann et al.
analyzed using methanolic extraction and liquid chromatography-
tandem mass spectrometry (LC-MS/MS) analysis as well as applying
alkaline hydrolysis and headspace-solid phase microextraction-
gas chromatography-mass spectrometry (HS-SPME-GC-MS) or
GC-MS analysis, respectively.
Materials and methods
Materials
Acetonitrile and methanol (MeOH) (both gradient grade) were
obtained from J.T. Baker (Deventer, the Netherlands) and Th.
Geyer (Renningen, Germany), n-hexane (p.a., ACS) was purchased
from Merck (Darmstadt, Germany). Formic acid (ROTIPURAN
98%, p.a.) and petroleum ether (ROTIPURAN 40-60C) were
obtained from Carl Roth (Karlsruhe, Germany) and acetone as
well as ethyl acetate (both p.a., ACS) from Sigma Aldrich
(Steinheim, Germany). Lecithin (egg phosphatidylcholine, as a
solubilizer for better stability of processed samples) was supplied
by Lipoid (Ludwigshafen, Germany). THC, CBN (1 mg/mL each),
and THC-D3 (0.1 mg/mL) were obtained from Cerilliant (Round
Rock, TX, USA). THCA-A and CBN-D3 (0.1 mg/mL) were purchased
from Lipomed (Arlesheim, Switzerland). THCA-A was dissolved in
MeOH (0.1 mg/mL). THCA-A-D3 was synthesized according to
Roth et al.[26] leading to a mixture of THCA-A-D3 and THC-D3
(>90% / ~9%). Deionized water was prepared using a cartridge
deionizer from Memtech (Moorenweis, Germany). Blank hair
was provided by volunteers and tested for the absence of canna-
binoids prior to use.
Sampling of hair
Hair samples from children (n = 41; age 7 months12 years), teen-
agers (n = 4; age 1317 years) and adults (n = 34; age: 1859
years) were collected between 2011 and 2013 on the order of
the Ofces of Social Services at the Senate of the Hanseatic City
of Bremen or at the Municipal Administrations of Bremerhaven
in the frame of a previously described project.[3] All the hair sam-
ples were taken from the posterior vertex region of the head as
close to the scalp as possible with a remaining hair length at
the scalp of approximately 12 mm and stored in the dark at
room temperature until analysis. The study was performed ac-
cording to the Helsinki ethical principles for medical research in-
volving human subjects of the Word Medical Association. Hair
sampling and analysis was performed either on the basis of a
written informed consent or on decision of the local family court.
In case of children the informed consent was signed by the par-
ents or caregivers. Emphasis was put on anonymous investiga-
tion in order to maintain condence with the families. The
samples were anonymized immediately after collection and only
the age of the individual was recognizable by the laboratory.
Only for some of the samples the connections between children
and adults were disclosed for extended interpretation.
Sample preparation
The sample preparation of the initial forensic analysis is described
in a previous paper[3] (Method 3). In this study, excessive hair
left over from this forensic investigation with a positive THC
result was reanalyzed. Thirty-two hair samples were shorter
than 7.0 cm and therefore processed without segmentation,
36 hair samples were segmented into one proximal segment
Drug Test. Analysis 2015, 7, 349357
 Drug Testing
Cannabinoid ndings in children hair what do they really tell us? and Analysis

of 6 cm length and the distal segment discarded. A further 11 hair Table 1. MRM transitions and corresponding voltages applied in
samples were not segmented despite a hair length between 7 and LC-MS/MS analysis (Method 1). (Q1) m/z of the precursor ion, (Q3)
20 cm as only a limited sample amount was still available. The m/z of the fragment ion, (DP) declustering potential, (CE) collision
energy, (CXP) collision cell exit potential.
samples or individual segments were washed by shaking for four
minutes with 4 mL water and twice with 4 mL acetone. Finally, the Analyte Q1, amu Q3, amu dwell DP, V CE, V CXP, V
hair was left to dry for 24 h. time, ms

THCA-A 357.2 313.2 20 -45 -34 -7

Method 1: Methanolic extraction and LC-MS/MS analysis 357.2 245.2 10 -45 -43 -5
THC 315.2 193.2 20 60 34 3
The washed hair samples were extracted and analyzed according 315.2 259.3 10 60 28 5
to a previously published method.[27] Briey, about 50 mg of the CBN 311.2 223.2 20 60 30 4
washed hair was weighted and cut into pieces of 12 mm length. 311.2 241.2 10 60 28 6
Extraction was carried out by addition of 2 mL MeOH and 20 L d3-THCA-A 360.3 316.4 20 -95 -34 -7
internal standard solution (3 ng THCA-A-D3, 20 ng THC-D3 and d3-THC 318.2 196.2 20 60 34 3
CBN-D3) (IS) and shaking for 4 h. After extraction and centrifuga- d3-CBN 314.3 223.2 20 60 30 4
tion at 2860 x g for 10 min (Heraeus Megafuge 1.0, Thermo
Scientic, Schwerte, Germany) 1.5 mL of the solution was transferred Bold: Ion transition used for quantication
into an empty LC-vial and evaporated to dryness under a stream of Entrance potential (EP): -10/10 V
nitrogen. The residue was reconstituted in 100 L acetonitrile
containing 0.1% formic acid and 0.25% lecithin and analyzed using
a fully validated LC-MS/MS method (lower limits of quantication 3 mL n-hexane:ethyl acetate (9: 1 v/v) was added and the sample
(LOQs): THCA-A 2.5 pg/mg, THC and CBN 20 pg/mg).[27] The was vortexed for 1 min. Finally the organic phase was transferred
LC-MS/MS system applied consisted of a Prominence HPLC sys- into a GC-vial and evaporated to dryness at 40 C under a
tem (two LC-20AD pumps, a SIL-20AC autosampler, a DGU- gentle stream of nitrogen. For GC-MS analysis the dried extracts
20A3 degasser, a CTO-20AC column oven and a CBM-20A con- were derivatized by the addition of 25 L N-methyl-N-
troller, Shimadzu, Duisburg, Germany) combined with a QTRAP trimethylsilyltriuoroacetamide (MSTFA) as well as 25 L ethyl
4000 triple quadrupole linear ion trap mass spectrometer tted acetate and incubated at 95 C for 30 min. After cooling, 1 L
with a TurboIon-Spray interface and Analyst software version was injected into the GC-MS system, consisting of a 6890 series
1.5.2 for data acquisition (AB Sciex, Darmstadt, Germany). Sep- GC system, a 5973 series mass selective detector, and a 7683 B
aration was performed applying gradient elution on a Luna series injector and using the software ChemStation G1701GA
C18 (2) column (150 x 2 mm, 5 m) with a corresponding version D.03.00.611 (Agilent, Waldbronn, Germany). The
guard column (C18 4 mm x 2 mm) (Phenomenex, conditions were as follows: splitless injection mode; column:
Aschaffenburg, Germany). Mobile phase A consisted of 0.1% HP-5MS (30 m x 0.25 mm I.D., 0.25 m lm thickness) (Agilent,
HCOOH in water and mobile phase B of 0.1% HCOOH in Waldbronn, Germany); injection port temperature: 250 C; carrier
ACN. The gradient started at 20% B for 1 min, increased to gas: helium; ow rate: 1.0 mL/min; oven temperature: initially
95% B in 7 min and was held at 95% for 4 min. Starting con- 140 C for 3 min, increased to 200 C at 30 /min, to 210 C at
ditions were restored within 1 min and the system was left to 5 C/min, to 240 C at 2 C/min, to 310 C at 30 C/min, 310 C
re-equilibrate at 20% B for 2.5 min prior to the injection of the for 4 min. The MS conditions were as follows: transfer line
next sample. The ow rate was set to 0.6 mL/min for the rst heater 280 C; ion source temperature 230 C; quadrupole
9.5 min, increased to 0.8 mL/min for 3.5 min and afterwards temperature 150 C; electron impact ionization (EI) mode, ion-
reduced to 0.6 mL/min for the rest of the run. The column ization energy 70 eV; electron multiplier voltage (EMV) 200 V
oven and the autosampler were heated to 50 C and 30 C, above autotune value. Analyses were performed in selected-ion
respectively, and the injection volume was 20 L. Data acqui- monitoring (SIM) mode using a validated method with the follow-
sition was performed in multiple reaction monitoring (MRM) ing settings: solvent delay 16.0 min; time window I, 16.020.5
mode including the MRM transitions listed in Table 1. The min, m/z 306, 374 (quantier (q)), 389 for THC-D3 and m/z 303,
mass spectrometer was operated using negative electrospray 371 (q), 386 for THC; time window II, 20.524.3 min, m/z 487 (q),
ionization (ESI) for THCA-A and its deuterated standard and posi- 488, 489, 502 for THCA A; time window III, 24.3028.33 min, m/z
tive ESI for THC, CBN and their respective deuterated standards. 374, 476, 491 for THC-COOH-D3.
The ion source temperature was set at 600 C, the ion source volt-
age at -4250 V / +4500 V, the curtain gas (N2) pressure was 35 psi,
Method 3: Alkaline hydrolysis and HS-SPME-GC-MS analysis
the ion source gas 1 (compressed air) 40 psi, ion source gas 2
(compressed air) 70 psi and the collision gas (N2) pressure was This method was used for the determination of THC, CBN, and
6 psi. The total cycle time was 1.66 s. CBD in the initial forensic analysis of the hair samples. For
alkaline hydrolysis only the proximal hair segments (06 cm)
or the full length in case of shorter hair were processed as
Method 2: Alkaline hydrolysis and GC-MS analysis
described by Nadulski et al.[23] In brief, between 15 and 30
Thirty-one of the 79 hair samples were retested using the method mg hair pieces were exactly weighed and digested with 1 mL
published by Auwrter et al.[17] For this purpose, 50 mg of the 1N NaOH containing each 10 ng D3-THC, D3-CBD and D3-CBN
washed hair was cut into 12 mm pieces. Afterwards, 1mL 1N as internal standards for 20 min at 80 C. The solution was
NaOH and 20l IS (D3-THC for THC, CBN and CBD; D3-THC-COOH two times extracted with 2 mL iso-octane, the solvent evapo-
for THCA-A; 5 ng each) were added and the samples were rated and the residue submitted to derivative headspace solid
351

hydrolyzed at 95C for 10 min. For liquid-liquid extraction (LLE) phase microextraction (HS-SPME) with N,O-bis(trimethylsilyl)

Drug Test. Analysis 2015, 7, 349357 Copyright 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/dta
 Drug Testing
and Analysis
triuoroacetamide (BSTFA) as derivatization agent and gas chro-
matography mass spectrometry in selected ion monitoring mode.
The method was validated according to international guidelines
with resulting LOQs of 0.01 ng/mg for all three compounds.
Analysis of marihuana and hashish samples
For marihuana samples, a representative amount was homoge-
nized using a knife mill and in case of the hashish samples, two
cores were drilled and the drilling chips were homogenized in a
mortar using liquid nitrogen. Afterwards, 10 mL internal standard
solution (0.5 mg/mL tribenzylamine in n-heptane) were added
into a glass vial containing 120 mg of hashish homogenate or
80400 mg of marihuana homogenate, respectively, and
sonicated at 50 C for 20 min. After the rst 10 min the sample
was shaken manually and additionally vortexed at the end of
the 20 min period. After deposition of the sediment, 500 L of
the supernatant were transferred into a GC vial and 200 l MSTFA
as well as 30 L pyridine were added. The closed vial was placed
in an oven at 70C for 20 min to perform silylation. Finally, the
samples were left to rest for 4 h at room temperature and after-
wards 1 L was injected into the gas-chromatography-ame ion-
ization detector (GC-FID) system (Focus GC or Trace GC Ultra,
both Thermo Scientic, Dreieich, Germany). The GC-FID condi-
tions were as follows: Split injection mode; column: BPX-5, SGE
(15 m x 0.32 mm ID, 0.25 m lm thickness); injection port
temperature: 250 C; detector (FID) port temperature 300 C; carrier
gas: helium; ow rate: 1.5 mL/min; oven temperature: initially 200C
for 2 min, increased to 235 C at 2 C/min, to 280 C at 50 C/min, different) and = 0.212 for THCA-A/THC (not signicantly
280 C for 4 min.
Statistical evaluations
Statistical analysis was performed using SPSS v19.0. Analysis
comparing drug concentration and concentration ratios uti-
lized analysis of variance (ANOVA) to determine whether or
not signicant differences existed in cannabinoid concentra-
tions and concentration ratios between children and adults
hair and between hair and marihuana or hashish. Signicance
was attributed to differences that attained a value of lower
than or equal to 0.05.
Results
In this study, hair samples from children and cannabis consuming
parents were analyzed by three different methods:
- Method 1: Extraction with methanol and LC-MS/MS. With this
method no decarboxylation of THCA-A to THC occurs and the
original concentrations of THC and THCA-A are measured.
- Method 2: Alkaline hydrolysis, liquid-liquid extraction, derivati-
zation and GC-MS. With this method, THCA-A was included in the
analysis and the extent of transformation can be checked.
- Method 3 (Initial forensic analysis): Alkaline hydrolysis, liquid-
liquid extraction and derivative headspace solid phase
microextraction with gas chromatography-mass spectrometry
(HS-SPME-GC-MS). With this method, THCA-A was not measured
but should be largely transformed to THC, as observed in the
studies conducted by Dussy et al.[28] and Auwrter et al.[17]
From the comparison between non-consuming children and
352
adults and between the results of the three methods further
wileyonlinelibrary.com/journal/dta Copyright 2014 John Wiley & Sons, Ltd.
B. Moosmann et al.
evidence about the incorporation mechanism into hair as a basis
for interpretation was expected.
Extraction with methanol and LC-MS/MS (Method 1)
A summary of the results obtained by applying Method 1
(methanol extraction, LC-MS/MS) can be found in Table 2, the
complete data are shown in the supplementary Table S1.
THCA-A could be detected in all but one hair sample and
ranged from 6.5 to 4700 pg/mg (median: 223 pg/mg). In 77
out of the 78 THCA-A positive cases, the concentration of
THCA-A was higher than the concentration of THC (median of
THCA-A/THC 4.2) and in 14 cases no THC could be detected
despite the presence of THCA-A (median of the THCA-A con-
centrations in these particular samples 63 pg/mg).
The results of a comparison between active consumers (adults,
age 1859 years) and passive exposure (children, age 12 years)
are shown in Table 2 and Figure 2. The four teenagers (age
1317 years) were excluded since the kind of contact with
cannabis is not clear. It can be seen from the mean values,
the medians and the box and whiskers plots that the concen-
trations of THCA-A as well as THC are clearly lower in hair of
the children than in hair of the adult consumers. However,
there is no signicant difference in the concentration ratio
THCA-A/THC between both groups if the specic relationship
between child and caregiver remains unconsidered in this
overall statistics. This is conrmed by analysis of variance
(ANOVA) with = 0.000 for both THCA-A and THC (signicantly
different).
However, the comparison within families leads to another
result. For this evaluation only a subset of 10 child-adult pairs
living in the same household could be included (Table 3), as
information about familiar connections was not disclosed in all
cases for data protection reasons. In these 9 families it can be
observed that in 9 of the 10 hair samples of children the ratio
THCA-A/THC is either higher than the ratio measured in the
samples from the cannabis consuming adults (n = 6), or only
THCA-A was detectable in the hair samples of the children
(n = 3). Analysis of variance (ANOVA) showed a signicant
difference between the two groups regarding the THCA-A/THC
ratios ( = 0.042; in case of absence of THC, the minimum ratio
based on the LOD of the method was used)
Cannabinoids and cannabinoid ratios in marihuana and hashish
In case of the consuming parents, no information was available
regarding the type of cannabis products used. As the investiga-
tion of conscated plant material shows a huge variation in the
THCA-A to THC ratios between different marihuana samples
and especially between marihuana and hashish samples (Table 4),
composition of the used plant material should have a strong
effect on the ratios found in hair. The comparison of the data
in Tables 2 and 4 and of the box an whiskers plots in Figure
3 show that the THCA-A/THC ratios in hair (all samples, mean
6.7, median 4.2) are in the range between hashish (mean 2.8,
median 2.1) and marihuana (mean 11.0, median 8.3). Statistical
comparison by ANOVA resulted in signicant differences for
hair samples and marihuana ( = 0.004 and F = 8.249), and
for hair samples and hashish ( = 0.000 and F = 35.321). It
can be seen from Figure 3 that the majority of the hair sam-
ples fall in the 2575 percentile range of the marihuana
Drug Test. Analysis 2015, 7, 349357
 Drug Testing
Cannabinoid ndings in children hair what do they really tell us? and Analysis

Table 2. Cannabinoid concentrations in hair measured by Method 1 and statistical comparison between children and adults. Samples of adoles-
cents (age 13 to 17 years, n = 4) were neither counted as adults nor as children.

All samples (n = 79) THCA-A, pg/mg THC, pg/mg CBN, pg/mg Ratio (n = 64*) THCA-A: THC

Mean 525 108 48 6.7

Median 223 35 11 4.2
SD 829 202 111 6.7
Range 04700 0 1200 0740 0.936
Children (n = 41) THCA-A, pg/mg THC, pg/mg CBN, pg/mg Ratio (n = 27*) THCA-A:THC
Mean 200 28 18 8.0
Median 99 13 0 4.3
SD 253 48 35 8.3
Range 01000 0250 0160 1.636

Adults (n = 34) THCA-A, pg/mg THC, pg/mg CBN, pg/mg Ratio (n = 33*) THCA-A:THC
Mean 961 214 89 5.7
Median 556 127 44 4.1
SD 1096 271 157 5.4
Range 464700 01200 0740 0.929
Comparison children vs. THCA-A, pg/mg THC, pg/mg CBN, pg/mg Ratio THCA-A:THC
adults by ANOVA
0.000 0.000 0.212
F 18.593 18.413 1.591
Different Yes Yes No

* Cases with THC < LOQ were not included

Figure 2. Box and whiskers plots of THCA-A and THC concentration as well as THCA-A to THC ratios in hair compared between adults and children
(n = 34 adults and 41 children), * and o indicate outliers.

samples whereas there is almost no overlap between the 2575
percentile ranges of hair and hashish samples.
Alkaline hair hydrolysis and GC-MS (Method 2)
A subset of the washed specimens, where enough material was
available as a left-over from Method 1 (30 samples), was analyzed
applying Method 2 in order to examine to which extent THCA-A
is transformed to THC under alkaline hydrolysis conditions. The
summarized results are shown in Table 5 and in detail in the sup-
plementary Table S2. It can be seen that despite 10 min treat-
an essential part of THCA-A remains undecomposed. Generally,
this can be an essential source of error in determination of THC
in hair by alkaline hydrolysis.
If THCA-A would be quantitatively transformed by decarboxyl-
ation to THC (Figure 1), and with consideration of the different
molecular masses, the expected total THC concentration can be
calculated according to Eqn (1).
THCtotal 314:47 x THCA-A= 358:47 THC (1)
The comparison in Table 5 and the Supplementary Table S2
353

ment in 1 N NaOH at 95 C and heating during derivatization shows that in 24 out of 30 hair samples these total THC

Drug Test. Analysis 2015, 7, 349357 Copyright 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/dta
 Drug Testing
and Analysis B. Moosmann et al.

Table 3. Comparison of THCA-A concentration, THC concentration concentrations calculated according to Eqn (1) for Method 1. It
and the ratio THCA-A/THC in hair obtained by Method 1 between chil- can be seen that the mean and the median of the concentrations
dren and related adult consumers living in the same household of THC (Method 3) and of calculated total THC (Method 1) are in
(households separated by dotted lines).
relatively good agreement. However, focusing on paired values
Case No. Age, years THCA-A, pg/mg THC, pg/mg THCA-A /THC measured in the samples of the same individual indicates a high
variation with the concentration ratio Method 1/Method 3 rang-
16 3 164 37 4.4
ing from 0.11 to 6.21.
66 31 2540 1227 2.1
Similar to Method 1, mean and median of this ratio are higher
11 2 301 70 4.3 than 1.00. However, on closer consideration of the individual
55 24 74 79 0.9 values in Table S3, this is the case only in 38 out of the 68 sam-
29 6 68 0 >6.8* ples. This includes 22 out of 32 children, two of the four teen-
59 26 128 34 3.8 agers and only 14 out of 32 adults. Possible reasons are
19 3 650 38 17.1 incomplete decarboxylation of THCA-A as well as degradation
37 10 73 0 >7.3* of both compounds under the more aggressive conditions.
70 34 613 161 3.8 In the case of CBN, Method 3 led to higher concentrations than
20 3 230 8 29.1 Method 1 in most of the samples (Table S3).
57 19 315 47 6.7
41 12 24 14 1.8
79 59 749 203 3.7 Discussion
22 4 448 40 11.3 A large variety of methods with respect to extraction, derivatiza-
63 29 452 90 5.0 tion, chromatography, and mass spectrometric detection is used
32 7 532 147 3.6 for the analysis of cannabinoids in hair samples in forensic case
71 34 652 260 2.5 work. Although results of prociency tests with spiked hair sam-
12 2 141 0 >14.1* ples indicate suitability of different methodological approaches,
53 23 472 99 4.8 the presence of THCA-A in authentic hair samples complicates
the situation. The presented data show that analytical results of
* Estimated using the LOD for THC (Method 1, 10 pg/mg) hair analysis for cannabinoids strongly depend on the applied
methodology mainly because of artifactual decarboxylation of
THCA-A under alkaline conditions and/or elevated temperatures.
concentrations from Method 1 (methanol extraction; LC-MS/MS)
This can lead to elevated THC concentrations but does not
were higher than the total THC concentration determined by
necessarily lead to a complete decarboxylation of THCA-A.
Method 2 (GC-MS after alkaline hydrolysis and LLE). The differ-
Nevertheless, only the THC concentration is generally used for
ence is more evident for children (mean ratio Method 1/Method
interpretation concerning patterns of cannabis use. Especially
2 1.64) than for adults (1.24). The smaller total concentration in
in cases applying alkaline hydrolysis and HS-SPME-GC-MS
Method 2 can be explained by a loss of both compounds caused
techniques it has to be carefully assessed how digestion time
by alkaline treatment and high GC injection temperature.[17,28]
and temperature inuence the degradation of the analytes
(e.g. degradation of THCA-A to THC, THC to CBN and CBN to
further oxidation products). On the other hand, remaining THCA-A
Alkaline hydrolysis and derivative HS-SPME-GC-MS (Method 3)
would lead to lower measured THC concentrations and could
The results of Method 1 were also compared with the concentra- give rise to false negative results in forensic cases.
tions of THC and CBN obtained in the initial forensic investigation The detection of THCA-A suggests that in almost all cases an
(given in Supplementary Table S3). For this comparison, only external contamination has occurred, since THCA-A is not incor-
samples were included where hair segments of equal lengths porated through the bloodstream in detectable amounts (LOD
were analyzed in both methods (n = 68). A summarized compari- in a previous study 50 pg/mg).[17] Furthermore, as side-stream
son is shown in Table 6 by using again the total THC marihuana smoke seems to play only a minor role as a source

Table 4. Cannabinoid content of seized marihuana and hashish samples.

Marihuana, n = 458 THC [%] THCA-A [%] Ratio THCA-A:THC CBD [%] CBDA-A [%] CBN [%]

Mean 1.1 6.7 11.0 0.0 0.1 0.0

Median 0.7 6.3 8.3 0.0 0.0 0.0
SD 1.4 4.8 11.6 0.2 0.3 0.1
Range 010.9 017.9 0119.5 03.3 03.5 01.3
Hashish, n = 180 THC [%] THCA-A [%] Ratio THCA-A:THC CBD [%] CBDA-A [%] CBN [%]
Mean 2.1 4.6 2.8 1.1 2.8 0.4
Median 2.0 4.1 2.1 0.9 3.0 0.2
SD 1.5 3.8 3.4 0.9 2.0 0.4
Range 09.3 015.5 030 05.7 08 02.1
354

wileyonlinelibrary.com/journal/dta Copyright 2014 John Wiley & Sons, Ltd. Drug Test. Analysis 2015, 7, 349357
 Drug Testing
Cannabinoid ndings in children hair what do they really tell us? and Analysis

At this point the question arises, if cut-offs could be

Figure 3. Box and whiskers plots of the concentration ratios THCA-A/
THC compared between 64 hair samples (children + teenager + adults),
458 marihuana samples and 180 hashish samples. For better visualization
5 marihuana outliers with THCA-A/THC ratios >55 were excluded from
this plot.
for THCA-A contamination of hair.[18] As a consequence, it
appears evident that some other form of external contamination
occurred in adults as well as in childrens hair (e.g. transfer
through contaminated hands or surfaces), and further studies
are encouraged for a denite proof of this assumption.
Regarding the interpretation of hair samples in particular from
young children, it has to be carefully evaluated if positive THC
ndings result from a smoke exposure or solely from external
contamination without smoke exposure. The higher THCA-A/
THC ratios in hair of children in comparison to their adult
caregivers within families indicate dominating external contami-
nation. In contrast, the ndings in the hair samples of the
cannabis consuming adults seem to result from a combination of ex-
ternal contamination through e.g. contaminated ngers, incorpora-
tion through side-stream smoke (only THC and CBN[18]) and from
the bloodstream, the latter both elevating the THC concentration.
established to differentiate the different types of contamination
and cannabis consumption. For this purpose, the ratio of
THCA-A to THC in the plant material has to be considered. Given
the large variability of this ratio in marihuana and hashish
samples which can be explained by different types of processing
and storage, it is clear that strict limits cannot serve as a valid
basis for interpretation. In addition, other aspects like progressive
decomposition of THCA-A to THC in the hair matrix as well as
efciency of incorporation and removal of the compounds have
to be considered.
On the other hand, in current forensic practice THC in hair is
quantitatively assessed using cut-off values. Different results from
different methods caused by varying transformation of THCA-A
to THC would mean disparity in interpretation. Therefore, harmo-
nization of analytical methods and a discussion on the necessity
of inclusion of THCA-A as an additional analyte are urgently re-
quired. Since, different from THC-COOH, a positive THC result is
generally interpreted as contact to or dealing with cannabis
products and not as a consumption, it would be reasonable to
determine total THC as a sum of individually determined THC
and THCA-A concentrations in order to obtain a more differenti-
ated picture.
In general, THCA-A could be a valuable marker facilitating the
interpretation of the results not only in child protection cases
but also in other issues such as hair samples from alleged canna-
bis growers or from law enforcement ofcers handling seized
plant material. For unbiased quantication of THCA-A next to
other cannabinoids, extraction in an organic solvent without
heating, sonication or alkaline hydrolysis, followed by either gas
chromatography after silylation or liquid chromatography with
mass spectrometric detection can be recommended. Analysis
for THCA-A may be of great value particularly in cases where
the sample amount does not allow for an additional analysis for
THC-COOH or when sufciently sensitive analytical instrumenta-
tion is not available.

Table 5. Concentrations of THCA-A and THC in hair obtained by Method 2 and comparison of the total THC concentrations calculated by Eqn (1)
between Methods 1 and 2.
All samples tested THCA-A Method 2, THC Method 2, Total THC conc. Method 1, Total THC conc. Method 2, Ratio Method 1:
with Method 1 pg/mg pg/mg pg/mg pg/mg Method 2
and Method 2
(n = 30)

Mean 65.8 373 622 445 1.48

Median 45.8 157 328 212 1.49
SD 82.4 615 915 679 0.56
Range 0419 02716 30 3956 192881 0.713.27
Children (n = 18) THCA-A Method 2, THC Method 2, Total THC conc. Method Total THC conc. Method Ratio Method 1:
pg/mg pg/mg 1, pg/mg 2, pg/mg Method 2
Mean 43.3 116 289 163 1.64
Median 29.6 65.4 152 112 1.59
SD 46.9 128 313 164 0.62
Range 0210 0536 301145 19720 0.813.27
Adults (n = 12) THCA-A Method 2, THC Method 2, Total THC conc. Method Total THC conc. Method Ratio Method 1:
pg/mg pg/mg 1, pg/mg 2, pg/mg Method 2
Mean 102 779 1122 868 1.24
Median 56.8 505 637 593 1.26
SD 113 844 1266 921 0.37
Range 18.8419 58.52716 653956 892881 0.711.92
355

Drug Test. Analysis 2015, 7, 349357 Copyright 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/dta
 Drug Testing
and Analysis B. Moosmann et al.

Table 6. Comparison of results obtained by method 1 with the results obtained by Method 3. The total THC concentration for Method 1 was
calculated from THC and THCA-A by Eqn (1).

All samples tested Total THC conc. THC conc. Total THC conc. CBN conc. CBN conc. CBN conc. Method 1:
with Method 1 Method 1, Method 3, Method 1: THC Method 1, Method 3, CBN conc.
and Method 3 pg/mg pg/mg conc. Method 3 pg/mg pg/mg Method 3
(n = 68, including
4 teenagers)

Mean 603 585 1.48 48.5 69.7 0.52

Median 214 195 1.14 10.7 30.0 0.41
SD 960 862 1.20 119 92.2 0.57
Range 5.75195 114330 0.116.21 0740 0370 02.75

Children (n = 32) Total THC conc. THC conc. Total THC conc. CBN conc. CBN conc. CBN conc. Method
Method 1, pg/mg Method 3, pg/mg Method 1: THC conc. Method 1, Method 3, 1: CBN conc.
Method 3 pg/mg pg/mg Method 3
Mean 183 133 1.92 11.0 34.6 0.27
Median 73.6 56.0 1.73 0.0 20.0 0.24
SD 230 251 1.38 23.1 66.4 0.29
Range 5.7874 111360 0.136.21 0120 0360 00.92
Adults (n = 32) Total THC conc. THC conc. Total THC conc. CBN conc. CBN conc. CBN conc. Method
Method 1, pg/mg Method 3, pg/mg Method 1: THC conc. Method 1, Method 3, 1: CBN conc.
Method 3 pg/mg pg/mg Method 3
Mean 1081 1093 1.10 91.8 110 0.60
Median 648 705 0.92 43.7 75.0 0.51
SD 1223 1018 0.90 162 101 0.68
Range 405195 1004330 0.114.77 0740 0370 02.75

Acknowledgements [12] C. Moore, S. Rana, C. Coulter, F. Feyerherm, H. Prest. Application of

The project was funded by the Deutsche Forschungsgemeinschaft
(DFG-AU: 324/3-1)
References
[1] P. Kintz, Analytical and Practical Aspects of Drug Testing in Hair. Taylor
& Francis Group, Boca Raton, FL, 2007.
[2] F.P. Smith, D.A. Kidwell. Cocaine in hair, saliva, skin swabs, and urine
of cocaine users children. Forensic Sci. Int. 1996, 83, 179.
[3] F. Pragst, S. Broecker, M. Hastedt, S. Herre, H. Andresen-Streichert,
H. Sachs, M. Tsokos. Methadone and illegal drugs in hair from children
with parents in maintenance treatment or suspected for drug abuse
in a German community. Ther. Drug Monit. 2013, 35, 737.
[4] D. Lewis, C. Moore, P. Morrissey, J. Leikin. Determination of drug
exposure using hair: Application to child protective cases. Forensic
Sci. Int. 1997, 84, 123.
[5] W. Schubert, V. Dittmann, J. Brenner-Hartmann. Urteilsbildung in der
Fahreignungsbegutachtung Beurteilungskriterien, Vol. 3, Kirschbaum
Verlag, Bonn, 2013.
[6] R. Kronstrand, I. Nystrm, M. Forsman, K. Kll. Hair analysis for drugs
in drivers license regranting. A Swedish pilot study. Forensic Sci. Int.
2010, 196, 55.
[7] B. Dufaux, R. Agius, T. Nadulski, H.-G. Kahl. Comparison of urine and
hair testing for drugs of abuse in the control of abstinence in drivers
license re-granting. Drug Test. Anal. 2012, 4, 415.
[8] SAMHSA. Vol. Federal Register, 69, no. 71, 19673-19732. US Depart-
ment of Health and Mental Services, Substance Abuse and Mental
Health Services Administration, Rockville, Maryland, 2004.
[9] M. Uhl, H. Sachs. Cannabinoids in hair: Strategy to prove marijuana/
hashish consumption. Forensic Sci. Int. 2004, 145, 143.
[10] D. Thieme, H. Sachs, M. Uhl. Proof of cannabis administration by
sensitive detection of 11-nor-Delta (9)-tetrahydrocannabinol-9-
carboxylic acid in hair using selective methylation and application
of liquid chromatography-tandem and multistage mass spectrome-
try. Drug Test. Anal. 2014, 6, 112.
[11] H. Sachs, U. Dressler. Detection of THCCOOH in hair by MSD-NCI
356
two-dimensional gas chromatography with electron capture chemi-
cal ionization mass spectrometry to the detection of 11-nor-9-
tetrahydrocannabinol-9-carboxylic acid (THC-COOH) in hair. J. Anal.
Toxicol. 2006, 30, 171.
[13] T. Mieczkowski. A research note: The outcome of GC-MS/MS conrmation
of hair assays on 93 cannabinoid (+) cases. Forensic Sci. Int. 1995, 70, 83.
[14] J.Y. Kim, J.C. Cheong, J.I. Lee, M.K. In. Improved gas chromatography-
negative ion chemical ionization tandem mass spectrometric
method for determination of 11-nor-9-tetrahydrocannabinol-9-
carboxylic acid in hair using mechanical pulverization and bead-
assisted liquid-liquid extraction. Forensic Sci. Int. 2011, 206, e99.
[15] M.A. Huestis, R.A. Gustafson, E.T. Moolchan, A. Barnes, J.A. Bourland,
S.A. Sweeney, E.F. Hayes, P.M. Carpenter, M.L. Smith. Cannabinoid
concentrations in hair from documented cannabis users. Forensic
Sci. Int. 2007, 169, 129.
[16] E. Han, H. Chung, J.M. Song. Segmental hair analysis for 11-Nor-9-
tetrahydrocannabinol-9-carboxylic acid and the patterns of cannabis
use. J. Anal. Toxicol. 2012, 36, 195.
[17] V. Auwrter, A. Wohlfarth, J. Traber, D. Thieme, W. Weinmann. Hair
analysis for 9-tetrahydrocannabinolic acid A--new insights into
the mechanism of drug incorporation of cannabinoids into hair.
Forensic Sci. Int. 2010, 196, 10.
[18] B. Moosmann, N. Roth, V. Auwrter. Hair analysis for THCA-A, THC
and CBN after passive in vivo exposure to marijuana smoke. Drug
Test. Anal. 2014, 6, 119.
[19] F. Pragst, M.A. Balikova. State of the art in hair analysis for detection
of drug and alcohol abuse. Clin. Chim. Acta 2006, 370, 17.
[20] G. Romano, N. Barbera, G. Spadaro, V. Valenti. Determination of
drugs of abuse in hair: Evaluation of external heroin contamination
and risk of false positives. Forensic Sci. Int. 2003, 131, 98.
[21] D. Wilkins, H. Haughey, E. Cone, M. Huestis, R. Foltz, D. Rollins.
Quantitative analysis of THC, 11-OH-THC, and THCCOOH in human
hair by negative ion chemical ionization mass spectrometry.
J. Anal. Toxicol. 1995, 19, 483.
[22] A. Salomone, E. Gerace, F. DUrso, D. Di Corcia, M. Vincenti. Simulta-
neous analysis of several synthetic cannabinoids, THC, CBD and CBN,
in hair by ultra-high performance liquid chromatography tandem
mass spectrometry. Method validation and application to real sam-

after HPLC clean-up. Forensic Sci. Int. 2000, 107, 239. ples. J. Mass Spectrom. 2012, 47, 604.

wileyonlinelibrary.com/journal/dta Copyright 2014 John Wiley & Sons, Ltd. Drug Test. Analysis 2015, 7, 349357

Cannabinoid ndings in children hair what do they really tell us?
[23] T. Nadulski, F. Pragst. Simple and sensitive determination of 9-
tetrahydrocannabinol, cannabidiol and cannabinol in hair by
combined silylation, headspace solid phase microextraction and gas
chromatography-mass spectrometry. J. Chromatogr. B 2007, 846, 78.
[24] F. Musshoff, H.P. Junker, D.W. Lachenmeier, L. Kroener, B. Madea. Fully
automated determination of cannabinoids in hair samples using head-
space solid-phase microextraction and gas chromatography-mass
spectrometry. J. Anal. Toxicol. 2002, 26, 554.
[25] European Monitoring Centre for Drugs and Drug Addiction. Preg-
nancy, childcare and the family: Key issues for Europes response to
drugs. Publications Ofce of the European Union, Luxembourg, 2012.
[26] N. Roth, A. Wohlfarth, M. Mller, V. Auwrter. Regioselective synthe-
sis of isotopically labeled 9-tetrahydrocannabinolic acid A (THCA-
A-D3) by reaction of 9-tetrahydrocannabinol-D3 with magnesium
methyl carbonate. Forensic Sci. Int. 2012, 222, 368.
Drug Test. Analysis 2015, 7, 349357 Copyright 2014 John Wiley & Sons, Ltd.
Drug Testing
and Analysis
[27] N. Roth, B. Moosmann, V. Auwrter. Development and validation of an
LC-MS/MS method for quantication of 9-tetrahydrocannabinolic acid
A (THCA-A), THC, CBN and CBD in hair. J. Mass Spectrom. 2013, 48, 227.
[28] F.E. Dussy, C. Hamberg, M. Luginbuhl, T. Schwerzmann, T.A.
Briellmann. Isolation of 9-THCA-A from hemp and analytical as-
pects concerning the determination of 9-THC in cannabis products.
Forensic Sci. Int. 2005, 149, 3.
Supporting information
Additional supporting information may be found in the online
version of this article at the publishers web site.
357
wileyonlinelibrary.com/journal/dta