Content uploaded by Joanne E.L. VanDerNagel
Author content
All content in this area was uploaded by Joanne E.L. VanDerNagel on Apr 24, 2016
Content may be subject to copyright.
ORIGINAL ARTICLE
Fast and Highly Selective LC-MS/MS Screening for THC and
16 Other Abused Drugs and Metabolites in Human Hair to
Monitor Patients for Drug Abuse
Remco A. Koster, BSc,*Jan-Willem C. Alffenaar, PhD, PharmD,*Ben Greijdanus, BSc,*
Joanneke E. L. VanDerNagel, MD,†‡ and Donald R. A. Uges, PhD, PharmD*
Background: To facilitate the monitoring of drug abuse by patients,
a method was developed and validated for the analysis of amphetamine,
methamphetamine, 3,4-methylenedioxymethamphetamine, methyle-
nedioxyamphetamine, methylenedioxyethylamphetamine, methyl-
phenidate, cocaine, benzoylecgonine, morphine, codeine, heroin,
6-monoacteylmorphine, methadone, 2-ethylidene-1,5-dimethyl-
3,3-diphenylpyrrolidine (EDDP), delta-9-tetrahydrocannabinol
(THC), nicotine, and cotinine in human hair.
Methods: The hair preparation method contains a 3-step wash
procedure with dichloromethane followed by a simultaneous hair
pulverization and extraction procedure with disposable metal balls.
The developed liquid chromatography tandem mass spectrometry
method uses a single injection to detect and confirm all 17 abused
drugs, including THC, within 4.8 minutes.
Results: Nicotine was validated with a linear range of 800–
25,000 pg/mg hair, and all other substances were validated with
a linear range of 30.0–2500 pg/mg hair. For inaccuracy and impreci-
sion, the overall bias did not exceed 28.2% and the overall coefficient
of variation did not exceed 17.7%. Autosampler stability was proven
for 48 hours at 108C for all substances. Analytical cutoff concentra-
tions were defined for each substance at the lowest validated
inaccuracy and imprecision concentration with a bias and coefficient
of variation within 15% and qualifier/quantifier ratios within 20% of
the set ratio. The analytical cutoff concentrations were 200 pg/mg for
codeine and 80.0 pg/mg for 6-MAM, heroin, EDDP, and THC. The
analytical cutoff concentration for nicotine was 800 pg/mg and for all
other validated substances 30.0 pg/mg. This method was successfully
applied to analyze hair samples from patients who were monitored for
drug abuse. Hair samples of 47 subjects (segmented into 129 samples)
showed 3,4-methylenedioxymethamphetamine, methylphenidate,
cocaine, benzoylecgonine, codeine, methadone, EDDP, THC, nico-
tine, and cotinine above the analytical cutoff.
Conclusions: The method was fully validated, including the
validation of the qualifier/quantifier ratios. The analysis of real hair
samples proved the efficacy of the developed method for monitoring
drug abuse. The results obtained by this method provide the physician
or health-care professional with extensive information about actual
drug abuse or relapse and can be used for patient-specific therapy.
Key Words: hair, drugs of abuse, LC-MS/MS, qualifier/quantifier
ratios
(Ther Drug Monit 2013;0:1–10)
INTRODUCTION
Examining human biological matrices for the presence
of drugs is an important task of forensic and clinical
toxicological laboratories. These analyses are mostly per-
formed for urine and blood and can provide information about
drug use over a period of days. Information about use over
longer periods of time can be provided by analyzing hair
samples, as those drugs are incorporated into the hair during
its growth. Because hair grows approximately 1 cm each
month, segmental analysis may distinguish single exposure
from long-term exposure.
1–4
Not only hair analysis can pro-
vide information about use over longer periods of time but
also its sample collection is preferable because it is less inva-
sive than obtaining venous blood samples and less embarrass-
ing and time consuming than supervised urine collection.
The mechanisms for drug incorporation into hair are
still under investigation, and various mechanisms are
described. Drugs are diffused from the blood capillaries into
the growing hair cells and are incorporated into the completed
hair shaft by sebum, sweat, and the surrounding tissues. It is
also possible that drugs are deposited on hair from the
external environment. Incorporation of drugs into hair is
dependent on 3 key factors, namely, basicity and lipophilicity
of the substance and melanin content of the hair. This results
in low incorporation rate of acidic substances into the hair.
For example, 11-nor-Δ
9
-tetrahydrocannabinol-11-carboxylic
acid, the metabolite of delta-9-tetrahydrocannabinol (THC),
and ritalinic acid, the metabolite of methylphenidate, are
only found in hair in extremely low concentrations.
1
The
Received for publication April 1, 2012; accepted July 3, 2013.
From the *Laboratory for Clinical and Forensic Toxicology and Drugs Analysis,
Department of Hospital and Clinical Pharmacy, University of Groningen,
University Medical Center Groningen, The Netherlands; †SumID-Project,
Zorgontwikkeling, Tactus Addiction Medicine, Deventer, The Netherlands;
and ‡ACSW-Nijmegen Institute for Scientist-Practitioners in Addiction,
Radboud University, Nijmegen, The Netherlands.
Supplemental digital content is available for this article. Direct URL citations
appear in the printed text and are provided in the HTML and PDF versions
of this article on the journal’s Web site (www.drug-monitoring.com).
The authors declare no funding or conflict of interest.
Correspondence: Remco A. Koster, BSc, Laboratory for Clinical and
Forensic Toxicology and Drugs Analysis, Department of Hospital and
Clinical Pharmacy, University of Groningen, University Medical Center
Groningen University of Groningen, PO Box 30.001, 9700 RB Gronin-
gen, The Netherlands (e-mail: r.koster@umcg.nl).
Copyright © 2013 by Lippincott Williams & Wilkins
Ther Drug Monit Volume 0, Number 0, Month 2013 1
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
incorporation rate of drugs into hair can vary between sub-
jects and makes hair analysis unsuitable for quantitative ther-
apeutic drug monitoring. On the other hand, hair analysis is
useful for monitoring patient compliance by detecting the
presence or absence of drugs in the hair.
1–3
To prepare a hair sample for analysis, a wash procedure
is applied followed by extraction of the substances. To improve
analytical performance, residues from cosmetics, sweat, and
sebum should be removed by washing the hair sample. A wash
procedure also helps avoid false-positive results by removing
possible passive external contamination of drugs from the
environment. Some methods described in the literature use no
wash procedure at all, whereas other methods use labor-
intensive wash procedures.
1,5,6
However, neither of these wash
procedures were able to provide enough distinction between
hair contamination with drugs and active drug use.
7,8
Dichloro-
methane (DCM) is considered to be a suitable wash solvent
because it does not swell the hair, in contrast to aqueous sol-
vents or methanol, and therefore, no relevant extraction will
take place during the wash procedure.
6
There are multiple ways to extract drugs from hair.
Previously described methods use soaking in extraction
solvent, sonication with methanol, dissolving under alkaline
or acidic conditions, or using enzymes to treat the hair
samples.
9–15
Dissolving hair under alkaline or acidic conditions
is a good way to increase the interaction surface, although these
conditions can cause the formation of metabolites from the
parent substances. It is known, for example, that benzoylecgo-
nine can be formed from cocaine under alkaline conditions.
6
Extraction by soaking or using enzymes is also very time
consuming.
5,10,15–17
To overcome these problems, a ball mill
was used to pulverize the hair sample and increase the inter-
action surface of the hair and the extraction solvent. This
method creates a more homogeneous sample, is fast, and does
not induce any chemical instability in the substances.
Gas chromatography–mass spectrometry (GC-MS) is
a well-established technique for the determination of abused
drugs and is widely used for hair analysis. GC-MS also
requires an extensive sample preparation like solid-phase
extraction and liquid–liquid extraction and often derivatiza-
tion of the analyzed substances.
1,8,18
For a sensitive LC-MS/
MS technique, there may be no need for extensive sample
preparation and derivatization, whereas a high selectivity is
accomplished through the selection of an ionized molecule
and usually the most intense fragment. To obtain a higher
level of selectivity, a second fragment can be monitored
simultaneously. The first fragment (quantifier) will be used
to calculate concentrations, whereas the second fragment
(qualifier) will be used for confirmation of the detected sub-
stance using a qualifier/quantifier ratio.
There are several parameters that have to be taken into
account for a hair sample to be considered positive. The last
wash should be analyzed to monitor possible external
contamination. The identity of the substance should be
confirmed by determining the ratio of the qualifier/quantifier
mass transitions. When possible, use of a drug should be
confirmed through the presence of its metabolites. The
concentration of an identified drug should exceed a cutoff
concentration. Cutoff concentrations for the presence of
a number of abuses in hair are provided by the Society of
Hair Testing (SoHT) and the Substance Abuse and Mental
Health Services Administration.
19,20
According to the world drug report of the United Nations
Office on Drugs and Crime, the most widely used drugs are
cannabis, amphetamines, heroin, and cocaine. The nonmedical
use of prescription drugs like methylphenidate or methadone is
also recognized as a growing problem. Consumption of
a combination of drugs has also become more common.
21
The aim of this study was to develop and validate
a hair preparation and analysis method for most widely
abused drugs and their metabolites to monitor patients with
a high risk of drug abuse. The following substances were
included in the method: amphetamine, methamphetamine,
3,4-methylenedioxymethamphetamine (MDMA), methylene-
dioxyamphetamine (MDA), methylenedioxyethylamphetamine
(MDEA), methylphenidate, cocaine, benzoylecgonine, mor-
phine, codeine, heroin, 6-monoacteylmorphine (6-MAM),
methadone, 2-ethylidene-1,5-dimethyl-3,3-diphenylpyrrolidine
(EDDP), THC, nicotine, and cotinine.
MATERIALS AND METHODS
Chemicals and Reagents
Separate reference solutions containing amphetamine,
methamphetamine, MDMA, MDA, MDEA, methylphenidate,
cocaine, benzoylecgonine, morphine, codeine, heroin, 6-MAM,
methadone, EDDP, THC, nicotine, and cotinine of 1.0 mg/mL
in methanol were used. Cotinine was purchased from Cerilliant
(Round Rock, TX). All other drugs were purchased from
Lipomed (Arlesheim, Switzerland). Deuterated internal stand-
ards (IS) were used for all drugs. Lipomed reference solutions
were used for 6-MAM-D
3
, benzoylecgonine-D
3
,cocaine-D
3
,
codeine-D
3
,EDDP-D
3
, heroin-D
3
,MDA-D
5
, MDEA-D
5
,
MDMA-D
3
,methadone-D
3
, methamphetamine-D
5
,morphine-
D
3
,andTHC-D
3
. Cerilliant reference solutions were used for
cotinine-D
3
,methylphenidate-D
9
,nicotine-D
4
, and amphet-
amine-D
10
. Analytical grade methanol, 25% ammonia, and
DCM were purchased from Merck (Darmstadt, Germany).
Purified water was prepared by a Milli-Q Integral System
(Millipore, Billerica, MA). Ammonium formate was purchased
from Acros (Geel, Belgium).
Equipment and Conditions
Polypropylene containers of 5 mL with screw caps
(Kartell; Noviglio, Italy) were used to weigh the hair in and to
perform the wash procedure and extraction. A mixer mill
(Retsch MM400, Haan, Germany) was used to wash, pulver-
ize, and extract the hair samples. Disposable 2-mm diameter
stainless steel balls (Retsch; Haan, Germany) were used to
pulverize the hair samples. Vortexing was performed with
a multi-tube vortexer (Labtek, Christchurch, New Zealand).
Mini-Uniprep syringeless filter vials with regenerated cellulose
media 0.2 mm pore size (Whatman, Kent, United Kingdom),
were used to filter the suspension after extraction.
All experiments were performed on an Agilent 6460A
(Santa Clara, CA) triple quadrupole LC-MS/MS system, with
an Agilent 1200 series combined LC system. The Agilent
Koster et al Ther Drug Monit Volume 0, Number 0, Month 2013
22013 Lippincott Williams & Wilkins
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
6460A mass-selective detector operated in heated electrospray-
positive ionization mode and performed dynamic multiple
reaction monitoring with unit mass resolution. High-purity
nitrogen was used for both the source and collision gas flows.
In the first quadrupole, single charged ions [M + H]
+
were
selected. All precursor ions, product ions, optimum fragmentor
voltage, collision energy values, and retention times are shown
in Table 1. The most intense daughter ion in the product spec-
trum was chosen as quantifier, and a second daughter ion was
chosen as qualifier. For methylphenidate, however, the frag-
mentation of the ionized molecule only resulted in one daugh-
ter ion in the mass spectrum. This has led to the use of only one
mass transition (quantifier) for methylphenidate. Peak height
ratios of the substance and its IS were used to calculate con-
centrations. For all substances, the capillary voltage was set at
4000 V, gas temperature at 3208C, gas flow at 13 L/min,
nebulizer gas at 60 psi, sheath gas temperature at 4008C, sheath
gas flow at 12 L/min, and the nozzle voltage at 0 V. The
Agilent 1200 series autosampler was set at 108C, and the inte-
grated column oven was set at a temperature of 408C. The
mobile phase consisted of methanol and 20 mmol/L ammo-
nium formate buffer, pH 7.0, containing 5% vol/vol methanol.
Analyses were performed on a 50 ·2.1-mm 5-mm HyPURITY
Aquastar analytical column from ThermoFisher Scientific
(Waltham, MA) equipped with a separate 0.5-mmfritfilter
(Varian, Palo Alto, CA). Chromatographic separation was per-
formed by means of a gradient with a flow of 0.5 mL/min and
a run time of 4.8 minutes. The gradient starts at 100%
20 mmol/L ammonium formate buffer, pH 7.0, with 5% vol/
vol methanol and increases in 3 minutes to 100% methanol,
a percentage to be maintained between 3 and 3.9 minutes. At
3.91 minutes, the gradient returns to 100% 20 mmol/L
TABLE 1. Mass Spectrometer Settings and Retention Times for All Substances
Substance
Precursor
Ion (m/z)
Product Ion
Quantifier (m/z)
Product Ion
Qualifier (m/z)
Fragmentor
Voltage (V)
CE
Quantifier
(V)
CE
Qualifier
(V)
Retention
Time (min)
Amphetamine 136.3 91.2 119.2 70 8 3 2.3
Amphetamine-D
10
146.2 98.2 NA 70 8 NA 2.3
Methamphetamine 150.2 91.1 119.2 80 12 5 2.7
Methamphetamine-D
5
155.3 91.1 NA 80 12 NA 2.7
MDMA 194.2 105.2 133.2 80 23 15 2.5
MDMA-D
3
197.3 105.1 NA 80 23 NA 2.5
MDA 180.2 105.2 133.2 70 20 15 2.2
MDA-D
5
185.3 110.2 NA 70 20 NA 2.2
MDEA 208.2 105.2 133.2 95 25 18 2.5
MDEA-D
5
213.4 105.1 NA 95 25 NA 2.5
Methylphenidate 234.2 84.2 —90 16 NA 2.8
Methylphenidate-D
9
243.4 93.2 NA 90 16 NA 2.8
Cocaine 304.2 182.2 150.2 130 15 23 2.9
Cocaine-D
3
307.3 185.2 NA 130 15 NA 2.9
Benzoylecgonine 290.2 168.2 105.1 120 15 30 2.3
Benzoylecgonine-D
3
293.2 171.2 NA 120 15 NA 2.3
Morphine 286.2 165.2 153.2 160 45 50 2.3
Morphine-D
3
289.3 165.2 NA 160 45 NA 2.3
Codeine 300.2 165.2 153.1 155 45 50 2.8
Codeine-D
3
303.3 165.2 NA 155 45 NA 2.8
6-MAM 328.2 165.3 211.2 120 45 20 2.6
6-MAM-D
3
331.2 165.2 NA 120 45 NA 2.6
Heroin 370.2 165.2 268.2 170 55 25 2.9
Heroin-D
3
373.4 165 NA 170 55 NA 2.9
Methadone 310.3 265.2 105.1 110 10 25 3.3
Methadone-D
3
313.3 268.2 NA 110 10 NA 3.3
EDDP 278.3 234.2 186.2 160 30 35 3.1
EDDP-D
3
281.3 234.2 NA 160 30 NA 3.1
Nicotine 163.2 130.1 106.2 100 16 10 2.9
Nicotine-D
4
167.3 134.1 NA 100 16 NA 2.9
Cotinine 177.2 80.2 98.1 140 20 14 2.1
Cotinine-D
3
180.2 80.2 NA 140 20 NA 2.1
THC 315.3 193.3 259.3 130 20 18 3.6
THC-D
3
318.4 196.3 NA 130 20 NA 3.6
CE, collision energy; NA, not applicable.
Ther Drug Monit Volume 0, Number 0, Month 2013 LC-MS/MS Screening to Monitor Drug Abused Patients
2013 Lippincott Williams & Wilkins 3
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
ammonium formate buffer, pH 7.0, with 5% methanol and
stabilizes until 4.8 minutes to prepare the chromatographic
system for the next injection. Agilent Masshunter software
for quantitative analysis (version B.04.00) was used to quantify
the analysis results.
Sample Preparation
Reference solutions of 1.0 mg/mL for each substance
were used to prepare 2 stock solutions in methanol containing
50 mg/L nicotine and 5 mg/L of all other substances. One set
of reference solution was used to prepare the stock solution
used for the calibration curve, and a separate set was used to
prepare the stock solution for the quality control samples. A
combined stock solution of 5 mg/L was also prepared for all
the deuterated IS. This combined IS stock solution was spiked
to the extraction solvent at 25 mcg/L, which consists of
analytical grade methanol. For the hair preparation, a lock of
hair was placed on a piece of clean paper. The paper was
folded around the hair sample, and markings were made to
distinguish the hair root and the tip of the hair. Every
centimeter of hair length was marked on the paper. An empty
5-mL container was tared at the analytical balance, and 1 cm
of the hair sample (50 mg) and paper were cut with a pair of
scissors above the container. The paper was removed from the
hair sample with a pair of tweezers in such a way that the hair
fell through the folded paper into the container without the
sample being touched by the tweezers or hands. When
necessary, another centimeter of the hair sample was cut to
achieve approximately 50 mg. Length and weight of the
sampled hair were noted. The hair sample was washed 3 times
with 3 mL DCM in a mixer mill at 30 Hz for 10 minutes. The
last wash was collected in a glass tube and evaporated to
dryness at ambient temperature under a nitrogen gas flow.
The residue of the last wash was reconstituted in 1.5 mL
extraction solvent by vortexing for 2 minutes. After complet-
ing the 3 wash steps, the hair samples were left for 15 minutes
to evaporate the residual DCM from the containers. Fifty
stainless steel balls and 1.5 mL extraction solvent were added,
and the samples were mixed at 30 Hz for 30 minutes until all
the hair was pulverized. After 30 minutes, the suspension was
transferred into a clean tube and centrifuged at 1780g. Of the
supernatant or the reconstituted last wash, 200 mL was trans-
ferred into a Whatman filter vial and 5 mL of the filtered
supernatant was injected into the LC-MS/MS system. Length
and weight of the hair samples were noted, and the concen-
trations were recalculated to agree with a 50-mg sample, as
used for the calibration curve, with the following formula:
concentration resulted from analysis
weighed amount of hair · 50.
Influence on Ion Suppression of Hair
Pulverization Followed by Extraction,
Compared With Simultaneous Hair
Pulverization and Extraction
To investigate the influence of ion suppression, 3
blank hair samples were prepared according to the described
extraction procedure. The hair pulverization was also
performed with 3 other identical blank hair samples. For
these hair samples, 1.5 mL extraction solution containing
25 mcg/L of the deuterated IS was added after pulverization
by metal balls, and the samples were shaken for 30 minutes.
All extracts and neat extraction solvent were spiked with
20,000 pg/mg hair for nicotine and 2000 pg/mg for all
other substances. Peak heights were compared to assess
ion suppression of the 2 different hair preparations.
The influence of ion suppression was calculated as
follows, where 100% represents no ion suppression:
Mean peak height of spiked extracts
Mean peak height of spiked neat solution · 100.
Method Validation
The performed validation included linearity, inaccuracy,
imprecision, selectivity, specificity, and stability. Extracts of
blank hair from one volunteer were collected for spiking the
calibration and validation samples. Using this extract, an
8-point calibration curve was spiked for all substances.
Nicotine was spiked at 800, 1500, 2500, 5000, 10,000,
20,000, and 25,000 pg/mg hair. The calibration curves for all
other substances were spiked at 30.0, 80.0, 150, 250, 500,
1000, 2000, and 2500 pg/mg hair. The inaccuracy and
imprecision were tested at the lower limit of quantification
(LLOQ) and at low, medium, and high concentrations. For the
assessment of inaccuracy and imprecision, concentrations were
800, 2000, 5000, and 20,000 pg/mg hair for nicotine and 30.0,
80.0, 200, and 2000 pg/mg for all other drugs. Validation was
performed with a maximum tolerated bias and coefficient of
variation (CV) of 20% at the LLOQ and 15% at all other
validation concentrations, including the stability validation. To
determine inaccuracy and imprecision, all concentrations were
measured 5-fold in 3 separate runs on separate days. For each
inaccuracy and imprecision, assessment bias and CV were
calculated per run. Within-run, between-run, and overall CVs
were calculated using 1-way analysis of variance. Eight
calibration points were used to determine linearity on
3 separate days. Stabilities of the substances were assessed in
the autosampler at 108C at low and high levels after 48 hours.
Selectivity and specificity were assessed by analyzing 6 blank
hair samples from different persons. Peaks found in the blank
hair samples should not exceed 20% of the peak height at
LLOQ level. During the method validation, the ion ratios for
the qualifier and the quantifier were evaluated at the LLOQ low
and medium levels of the inaccuracy and imprecision study. All
ratios of the qualifier and the quantifier found during the inac-
curacy and imprecision validation were required to be within
20% of the ratio set in each validation run. Analytical cutoff
concentrations were defined for each substance at the lowest
concentration validated for inaccuracy and imprecision with the
bias and CV within 15% and all qualifier/quantifier ratios within
20% of the set ratio during the whole validation. Carryover was
monitored during validation, where the peak height of the first
blank after the highest calibrator should not exceed 20% of the
LLOQ. Performing a true recovery assessment with a hair
matrix is analytically impossible because drugs need to fully
incorporate into the hair. Bathing the hair in a drug-containing
solution would provide no quantifiable recovery. To provide
information about the consistency of the recovery with a real
hair sample, the imprecision of the developed method was
Koster et al Ther Drug Monit Volume 0, Number 0, Month 2013
42013 Lippincott Williams & Wilkins
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
further determined by the analysis of five 50-mg hair samples
obtained from a homogeneous positive hair sample for THC.
The hair sample resulted from a general haircut and consisted of
short pieces with a maximum length of 0.5 cm.
Routine Analysis
For purposes of routine analysis, hair samples were
collected from patients with a mild or borderline intellectual
disability. These patients were under institutional care for
their disability. The hair analysis was used to validate
a questionnaire for drug abuse in this target group.
22,23
All
patients (and, if applicable, their guardians) gave informed
consent to participate in the study, which was approved by
a certified medical ethical committee.
The collected hair samples were taken from the
posterior vertex and varied from short pieces of cut hair to
hair strings with a maximum length of 37 cm. When possible,
the hair strings were segmented to reach 50 mg per sample.
The results were corrected for the weighed amount of hair per
sample. For patient monitoring, the validated analytical cutoff
concentrations were used to determine drug abuse instead of
those proposed by the SoHT or Substance Abuse and Mental
Health Services Administration.
RESULTS
Influence on Ion Suppression of Hair
Pulverization Followed by Extraction,
Compared With Simultaneous Hair
Pulverization and Extraction
Table 2 shows the influence of ion suppression for hair
pulverization followed by extraction and simultaneous hair
pulverization and extraction for all substances and their deu-
terated IS. For hair pulverization followed by extraction, the
remaining peak heights caused by ion suppression ranged
from 41% to 78% with a mean of 57% (611.1) for all sub-
stances and 41% to 74% with a mean of 55% (610.6) for all
deuterated IS. For simultaneous hair pulverization and extrac-
tion, the remaining peak heights caused by ion suppression
ranged from 45% to 95% for all substances with a mean of
78% (613.1). The remaining peak heights for the deuterated
IS ranged from 41% to 89% with a mean of 73% (613.5).
Almost all substances and their deuterated IS were equally
affected by ion suppression. Deviations of less than 10%
between each substance and its deuterated IS were observed
for both extractions. Simultaneous hair pulverization and
extraction showed a larger difference of 17% less ion sup-
pression for morphine compared with morphine-D
3
. Com-
pared with hair pulverization followed by extraction,
simultaneous pulverization and extraction showed, respec-
tively, 21% and 18% less average ion suppression for all
substances and their deuterated IS and contributed to an effi-
cient hair extraction procedure.
Method Validation
Table 3 shows the validation results for linearity, inac-
curacy, imprecision, and stability. Nicotine was validated at
a linear range of 800–25,000 pg/mg. All other substances were
validated at a linear range of 30.0–2500 pg/mg hair. For inac-
curacy and imprecision, the highest overall bias found during
the validation was 28.2% at the LLOQ of heroin (30.0 pg/mg),
whereas the highest overall CV was 17.7% at the LLOQ of
THC (30.0 pg/mg). Autosampler stability was proven for 48
hours at 108C for all substances with a maximum overall bias of
4.8%. Selectivity and specificity showed no interfering peaks of
more than 20% of the LLOQ. In each validation run, the ion
ratio of the qualifier and quantifier was determined for each
substance. The analytical cutoff concentrations were defined
as the lowest concentrations with the bias and CV within
15% and the qualifier/quantifier ratios within 20% of the mean
ion ratio found during validation. Mean ion ratios and their
maximum deviations found during the validation are shown
in Table 4. It can be seen that 6-MAM, heroin, EDDP, THC,
TABLE 2. Influence on Ion Suppression of Hair Pulverization
Followed by Extraction, Compared With Simultaneous Hair
Pulverization and Extraction, Where 100% Represents No Ion
Suppression
Substance
Ion Suppression
Pulverization With
Methanol (%)
Ion Suppression
Pulverization Without
Methanol (%)
Amphetamine 82 46
Amphetamine-D
10
86 48
Methamphetamine 82 57
Methamphetamine-D
5
81 57
MDA 77 42
MDA-D
5
75 40
MDEA 85 62
MDEA-D
5
83 60
MDMA 85 62
MDMA-D
3
80 60
Methylphenidate 87 70
Methylphenidate-D
9
84 70
Cocaine 91 78
Cocaine-D
3
88 74
Benzoylecgonine 94 73
Benzoylecgonine-D
3
85 65
Morphine 78 44
Morphine-D
3
61 41
Codeine 64 55
Codeine-D
3
58 51
6-MAM 76 49
6-MAM-D
3
67 47
Heroin 72 59
Heroin-D
3
70 58
Methadone 54 51
Methadone-D
3
55 51
EDDP 81 71
EDDP-D
3
80 70
Nicotine 69 50
Nicotine-D
4
61 44
Cotinine 95 65
Cotinine-D
3
89 61
THC 45 41
THC-D
3
41 41
Ther Drug Monit Volume 0, Number 0, Month 2013 LC-MS/MS Screening to Monitor Drug Abused Patients
2013 Lippincott Williams & Wilkins 5
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
TABLE 3. Validation Results of All Substances
Substance
Correlation Coefficient
(Linear Range) (pg/mg)
Concentration
(pg/mg)
Within Run CV
(%)
Between Run CV
(%)
Overall CV
(%)
Overall Bias
(%)
Amphetamine 0.99837 (30–2500) LLOQ (30.0) 7.9 0.0 7.9 27.0
Low (80.0) 4.9 4.7 6.7 22.2
Medium (200) 3.6 2.6 4.5 3.5
High (2000) 4.8 2.8 5.6 2.3
AS stab low 48 h 2.4 NA NA 4.8
AS stab high 48 h 3.6 NA NA 0.8
Methamphetamine 0.99885 (30–2500) LLOQ (30.0) 2.6 5.7 6.3 0.0
Low (80.0) 1.6 0.7 1.8 24.3
Medium (200) 1.8 1.6 2.4 22.5
High (2000) 1.2 2.3 2.6 23.3
AS stab low 48 h 0.9 NA NA 20.6
AS stab high 48 h 1.3 NA NA 21.1
MDMA 0.99843 (30–2500) LLOQ (30.0) 4.4 1.8 4.7 3.2
Low (80.0) 3.9 2.3 4.5 23.4
Medium (200) 2.3 2.7 3.6 20.8
High (2000) 0.9 3.7 3.8 20.4
AS stab low 48 h 2.5 NA NA 0.5
AS stab high 48 h 1.5 NA NA 1.1
MDA 0.99868 (30–2500) LLOQ (30.0) 2.8 7.5 8.0 5.9
Low (80.0) 2.1 3.5 4.1 23.4
Medium (200) 2.4 1.1 2.7 22.1
High (2000) 1.5 4.8 5.0 21.2
AS stab low 48 h 1.2 NA NA 20.8
AS stab high 48 h 0.9 NA NA 0.2
MDEA 0.99481 (30–2500) LLOQ (30.0) 4.7 5.7 7.4 4.1
Low (80.0) 3.5 1.1 3.7 20.3
Medium (200) 1.7 1.4 2.2 0.6
High (2000) 1.1 2.2 2.5 20.8
AS stab low 48 h 1.5 NA NA 0.3
AS stab high 48 h 1.0 NA NA 20.2
Methylphenidate 0.99911 (30–2500) LLOQ (30.0) 2.8 7.8 8.3 1.1
Low (80.0) 2.6 0.0 2.6 23.1
Medium (200) 1.9 1.2 2.2 22.7
High (2000) 2.7 5.1 5.8 24.3
AS stab low 48 h 3.0 NA NA 21.9
AS stab high 48 h 1.8 NA NA 23.5
Cocaine 0.99812 (30–2500) LLOQ (30.0) 4.7 1.8 5.1 4.3
Low (80.0) 2.0 0.0 2.0 22.8
Medium (200) 1.5 4.0 4.3 22.3
High (2000) 1.0 3.4 3.5 23.7
AS stab low 48 h 0.8 NA NA 22.0
AS stab high 48 h 0.9 NA NA 21.9
Benzoylecgonine 0.99807 (30–2500) LLOQ (30.0) 2.6 3.9 4.7 4.8
Low (80.0) 1.4 0.0 1.4 21.9
Medium (200) 1.0 1.7 2.0 20.9
High (2000) 0.6 3.3 3.4 21.6
AS stab low 48 h 1.1 NA NA 0.4
AS stab high 48 h 0.7 NA NA 0.3
Morphine 0.99875 (30–2500) LLOQ (30.0) 5.1 5.0 7.1 21.0
Low (80.0) 4.6 1.0 4.7 26.2
Medium (200) 2.5 2.0 3.2 23.0
High (2000) 2.4 1.3 2.7 20.4
AS stab low 48 h 2.8 NA NA 22.3
AS stab high 48 h 1.3 NA NA 1.9
Koster et al Ther Drug Monit Volume 0, Number 0, Month 2013
62013 Lippincott Williams & Wilkins
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
TABLE 3. (Continued) Validation Results of All Substances
Substance
Correlation Coefficient
(Linear Range) (pg/mg)
Concentration
(pg/mg)
Within Run CV
(%)
Between Run CV
(%)
Overall CV
(%)
Overall Bias
(%)
Codeine 0.99843 (30–2500) LLOQ (30.0) 7.6 10.9 13.3 23.2
Low (80.0) 5.9 2.1 6.2 22.7
Medium (200) 4.5 3.7 5.9 1.1
High (2000) 2.8 1.2 3.1 0.7
AS stab low 48 h 4.1 NA NA 1.2
AS stab high 48 h 1.9 NA NA 0.8
6-MAM 0.99813 (30–2500) LLOQ (30.0) 8.6 1.4 8.7 22.3
Low (80.0) 2.9 2.0 3.5 25.0
Medium (200) 3.2 2.6 4.1 22.3
High (2000) 1.5 2.4 2.8 20.9
AS stab low 48 h 2.5 NA NA 24.2
AS stab high 48 h 1.7 NA NA 20.1
Heroin 0.99867 (30–2500) LLOQ (30.0) 8.3 0.0 8.3 28.2
Low (80.0) 7.2 3.1 7.9 24.7
Medium (200) 4.2 1.4 4.4 24.1
High (2000) 1.6 1.1 2.0 21.4
AS stab low 48 h 4.0 NA NA 24.8
AS stab high 48 h 1.6 NA NA 20.8
Methadone 0.99851 (30–2500) LLOQ (30.0) 1.3 5.4 5.5 2.5
Low (80.0) 1.9 0.6 2.0 22.6
Medium (200) 0.8 3.0 3.1 21.3
High (2000) 1.0 3.7 3.9 21.8
AS stab low 48 h 1.1 NA NA 20.5
AS stab high 48 h 0.8 NA NA 0.6
EDDP 0.99842 (30–2500) LLOQ (30.0) 3.9 0.0 3.9 0.4
Low (80.0) 2.1 0.9 2.3 26.3
Medium (500) 1.4 1.6 2.2 22.3
High (2000) 0.6 2.4 2.4 21.2
AS stab low 48 h 1.2 NA NA 21.3
AS stab high 48 h 0.5 NA NA 0.5
Nicotine 0.99605 (800–25,000) LLOQ (800) 8.3 2.1 8.6 22.2
Low (2000) 4.2 0.0 4.2 22.5
Medium (5000) 3.8 4.0 5.5 20.5
High (20,000) 2.6 3.0 4.0 1.3
AS stab low 48 h 3.5 NA NA 22.9
AS stab high 48 h 3.5 NA NA 1.5
Cotinine 0.99847 (30–2500) LLOQ (30.0) 9.0 10.9 14.1 24.3
Low (80.0) 3.5 2.6 4.3 23.2
Medium (200) 2.3 1.6 2.8 21.7
High (2000) 2.0 2.8 3.5 22.8
AS stab low 48 h 1.0 NA NA 20.1
AS stab high 48 h 1.4 NA NA 20.3
THC 0.99423 (30–2500) LLOQ (30.0) 12.3 12.8 17.7 21.2
Low (80.0) 9.7 0.0 9.7 0.8
Medium (200) 3.7 0.0 3.7 21.3
High (2000) 1.8 3.2 3.7 0.5
AS stab low 48 h 4.1 NA NA 22.1
AS stab high 48 h 1.3 NA NA 4.8
LLOQ, low, medium, and high are the validation concentrations used for inaccuracy and imprecision assessment; AS stab low/high 48 hours represents stabil ity in the autosampler
at 108C for 48 hours after sample preparation.
NA, not applicable.
Ther Drug Monit Volume 0, Number 0, Month 2013 LC-MS/MS Screening to Monitor Drug Abused Patients
2013 Lippincott Williams & Wilkins 7
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
and codeine have more than 20% deviation from their mean ion
ratio at the LLOQ level and acceptable deviations at the low
level, except for codeine. Codeine showed to have a 21.2%
deviation at the low level and an acceptable maximum devia-
tion of 6.6% at the medium level. All other substances dis-
played deviations within 20% from the mean ion ratio at
the LLOQ level (see Figure, Supplement Digital content 1,
http://links.lww.com/TDM/A57). The analytical cutoff concen-
trations based on the analytical performance and the screening
cutoff concentrations set by the SoHT are shown in Table 5.
For 11 substances, the screening cutoff concentrations are set
by the SoHT, whereas for benzoylecgonine and EDDP, only
confirmatory cutoff concentrations of 50 pg/mg are provided.
For 4 substances, there are no cutoff concentrations set by the
SoHT.
24
Because of the lower sensitivity of the qualifier, the
qualifier/quantifier ratios set the analytical cutoff concentrations
to 200 pg/mg for codeine and 80 pg/mg for 6-MAM, heroin,
EDDP, and THC. The analytical cutoff concentration for nico-
tine is 800 pg/mg, and for all other validated substances, the
method cutoff concentration is 30.0 pg/mg. All analytical cutoff
concentrations are lower than or equal to the screening cutoff
concentrations set by the SoHT. The analytical cutoff concen-
tration of 30 pg/mg for benzoylecgonine is even lower than
the confirmatory cutoff of 50 pg/mg set by the SoHT. For
EDDP, the analytical cutoff of 80 pg/mg is higher than the
confirmatory cutoff of 50 pg/mg set by the SoHT. Because
the analytical cutoff of 80 pg/mg is close to the confirmatory
cutoff of 50 pg/mg set by the SoHT, it is considered to be
adequate for screening purposes. No carryover was detected
during method development and validation. To determine
the imprecision of the method for THC, the results of the
5 hair extracts from subject 11 showed reproducible THC
concentrations and thus consistent recoveries, with a mean con-
centration of 348 (627.1) pg/mg and a CV of 7.8% (Table 6).
Routine Analysis
The developed analytical method was used to analyze the
hair of 47 subjects; hair strings were segmented when possible.
This resulted in the analysis of 129 hair extracts. During the
routine analysis, MDMA, methylphenidate, cocaine, benzoy-
lecgonine, codeine, methadone, EDDP, nicotine, and cotinine
were detected above analytical cutoff concentrations in the hair
samples. An acceptable ratio of the qualifier/quantifier mass
transition was also found for each positive sample. Eighty
extracts were tested positive (.800 pg/mg hair) for the presence
of nicotine, 59 of which were confirmed by the presence of
cotinine (.30.0 pg/mg hair). The concentrations found for nic-
otine ranged from 816 to 176,387 pg/mg and for cotinine from
32 to 10,296 pg/mg hair. The other detected substances are
shown in Table 6. For subjects 3 and 5, cocaine and methadone
were present in the last wash solution. The concentrations found
in the corresponding hair extracts were 20–78 times higher than
those found in the last wash. In addition, the metabolites ben-
zoylecgonine and EDDP showed to be present in the extracts
and not in the last wash solution. This indicates that the wash
procedure does not remove intrinsic drugs from the hair sample
and that the outside of the hair may have been contaminated
with the used drugs. Nevertheless, the combination of these
results leads to the conclusion that cocaine and methadone were
used by these subjects. The hair samples of subjects 6 and 7
showed the presence of methylphenidate. This led to the con-
clusion that these subjects had used methylphenidate, either as
TABLE 4. Mean Qualifier/Quantifier Ratios During the
Validation and the Maximum Deviation Found at the LLOQ,
Low, and Medium Levels
Substance Mean Ion Ratio
Maximum Deviation (%)
LLOQ Low Medium
Amphetamine 70.9 5.3 2.5 3.5
Methamphetamine 29.8 12.6 5.3 6.0
MDMA 60.1 17.0 7.4 6.2
MDA 69.2 13.2 9.3 5.2
MDEA 61.2 19.2 12.4 7.2
Methylphenidate NA NA NA NA
Cocaine 8.4 10.8 5.4 4.6
Benzoylecgonine 30.4 9.2 4.5 2.8
Morphine 87.2 17.8 11.4 14.4
Codeine 85.0 39.1 21.2 6.6
6-MAM 61.0 21.4 11.8 7.9
Heroin 64.5 21.8 16.9 12.2
Methadone 49.2 6.3 6.0 3.1
EDDP 30.7 21.4 9.7 4.4
Nicotine 45.4 4.4 4.0 4.0
Cotinine 23.1 6.8 5.5 3.1
THC 38.2 79.1 17.1 12.4
NA, not applicable.
TABLE 5. Validated Analytical Cutoff Concentrations Based on
the Performance of the Quantifier and Qualifier and the
Screening Cutoff Concentrations Set by the SoHT, Where
Available
Substance
Method Cutoff
Hair (pg/mg)
SoHT Cutoff
Hair (pg/mg)
Amphetamine 30 200
Methamphetamine 30 200
MDMA 30 200
MDA 30 200
MDEA 30 200
Methylphenidate 30 —
Cocaine 30 500
Benzoylecgonine 30 50*
Morphine 30 200
Codeine 200 200
6-MAM 80 200
Heroin 80 —
Methadone 30 200
EDDP 80 50*
Nicotine 800 —
Cotinine 30 —
THC 80 100
*No screening cutoff concentration available, confirmation cutoff concentration
used.
Koster et al Ther Drug Monit Volume 0, Number 0, Month 2013
82013 Lippincott Williams & Wilkins
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
prescription drug or as abused drug. Table 6 also shows that
segmental analysis proved repeated use of methylphenidate,
cocaine, and codeine.
DISCUSSION
This LC-MS/MS method was developed for fast
screening of abused drugs in human hair in patients who
were monitored for drug abuse. The developed method used
a fast sample preparation with a ball mill to simultaneously
pulverize and extract the hair sample. False-positive results
are minimized with the use of a wash procedure. The
developed method was fully validated with enhanced selec-
tivity by validated qualifier/quantifier ratios.
The ion suppression tests of simultaneous hair pulver-
ization and extraction showed 17% less ion suppression for
morphine compared with morphine-D
3
. The observed ion
suppression differences for all other substances and their deu-
terated IS are no more than 9% and are able to correct well
for signal changes because of matrix effects. Even though
morphine-D
3
corrected less than expected for signal changes
because of matrix effects, it is still considered to be the best
available IS and proved suitable for the developed and vali-
dated screening method.
The developed simultaneous pulverization and extraction
method is very time efficient compared with separate pulver-
ization followed by extraction and showed less ion suppression
for all substances and their deuterated IS. These 2 benefits
contributed to a fast and efficient hair extraction procedure.
Several methods have been described for the analysis
of multiple substances in hair. Many methods describe an
extensive sample preparation by incubating the hair sample
followed by liquid–liquid extraction and/or solid-phase
extraction and analysis using GC-MS or LC-MS/MS.
25,26
Despite extensive sample preparations, these methods show
comparable LLOQs. Previously published methods describ-
ing the analysis of multiple drugs often do not include THC in
the same analysis.
5,27–30
And despite the fact that many pub-
lished methods are extensively validated, no clear restrictions
are applied for the validation of qualifier mass transitions for
LC-MS/MS analysis. Several methods describe identification
criteria for ion ratios found during analysis, but validated
results for ion ratios are never clearly mentioned.
5,15,26,28–30
Considering the fact that the quantifier mass transition is
almost always the most intense fragment and the qualifier
mass transition the second most intense, it is imperative that
the latter be proven to be reliable at the validated cutoff
concentration. Validation of the qualifier/quantifier ratios
have resulted in higher analytical cutoff concentrations for
codeine, 6-MAM, heroin, EDDP, and THC than the validated
LLOQs for inaccuracy and imprecision of these quantifiers.
This shows that although the quantifier provides good inac-
curacy and imprecision, below these analytical cutoff concen-
trations, the inaccuracy of the ion ratio may be insufficient for
confirmation of the detected substance. Although this raises
the analytical cutoff concentrations for these substances, they
are still within the screening cutoff concentrations set by the
SoHT.
24
This not only shows the usefulness of validating the
qualifier/quantifier ratios but also underlines the selectivity
and reliability of the method described here.
The routine analysis showed that multiple substances
were detected in the hair above the analytical cutoff. In some
hair samples, the parent drug was also detected in the last wash.
Although the concentrations found in the extracts were 20–78
times higher than those found in the last wash, the presence of
the metabolites definitely proved use of these drugs.
CONCLUSIONS
The developed method can detect and confirm 17
substances, including THC, within one analytical run of
4.8 minutes. With this method, patients can be monitored for
drug abuse without invasive sample collection. Ongoing
TABLE 6. The Results of the Analysis of Human Hair Samples
Subject
Hair Length
Segment (cm) Substance
Last
Wash
(pg/mg)
Hair
Extract
(pg/mg)
11–2 MDMA ,30 44
20–6 MDMA ,30 165
31–7 Benzoylecgonine ,30 40
0–2 Cocaine ,30 308
Benzoylecgonine ,30 61
2–5 Cocaine ,30 601
Benzoylecgonine ,30 82
5–8 Cocaine 70 1420
Benzoylecgonine ,30 183
8–11 Cocaine 77 1783
Benzoylecgonine ,30 198
15–22 Cocaine 88 3891
Benzoylecgonine ,30 193
40–7 Cocaine ,30 1235
50–10 Cocaine 497 38,762
Benzoylecgonine ,30 6052
Methadone 1148 45,402
EDDP ,80 1573
60–2 Methylphenidate ,30 281
2–3 Methylphenidate ,30 174
70–2 Methylphenidate ,30 199
2–4 Methylphenidate ,30 159
4–7 Methylphenidate ,30 129
80–1 Codeine ,30 258
1–2 Codeine ,30 289
90–2 THC ,80 139
10 0–6 THC ,80 ,80
6–10 THC ,80 ,80
10–14 THC ,80 ,80
18–22 THC ,80 89
28–37 THC ,80 121
11 0–0.5 THC ,80 356
0–0.5 THC ,80 385
0–0.5 THC ,80 337
0–0.5 THC ,80 303
0–0.5 THC ,80 358
The start of the "hair length segment" is the distance of the segment measured from
the scalp.
Ther Drug Monit Volume 0, Number 0, Month 2013 LC-MS/MS Screening to Monitor Drug Abused Patients
2013 Lippincott Williams & Wilkins 9
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
routine analysis proved the developed method to be a reliable
tool for determining 17 abused drugs and metabolites in
human hair samples from patients who were monitored for
drug abuse. The results provide the physician or health-care
professional with extensive information about drug abuse or
relapse in drug abuse and can be used for patient-specific
therapy, thus improving patient care.
REFERENCES
1. Pragst F, Balikova MA. State of the art in hair analysis for detection of
drug and alcohol abuse. Clin Chim Acta. 2006;370:17–49.
2. Kintz P, Marescaux C, Mangin P. Testing human hair for carbamazepine
in epileptic patients: is hair investigation suitable for drug monitoring?
Hum Exp Toxicol. 1995;14:812–815.
3. Takiguchi Y, Ishihara R, Torii M, et al. Hair analysis of flecainide for
assessing the individual drug-taking behavior. Eur J Clin Pharmacol.
2002;58:99–101.
4. Sato H, Uematsu T, Yamada K, et al. Chlorpromazine in human scalp
hair as an index of dosage history: comparison with simultaneously
measured haloperidol. Eur J Clin Pharmacol. 1993;44:439–444.
5. Kronstrand R, Nystrom I, Strandberg J, et al. Screening for drugs of abuse
in hair with ion spray LC-MS-MS. Forensic Sci Int. 2004;145:183–190.
6. Musshoff F, Madea B. Analytical pitfalls in hair testing. Anal Bioanal
Chem. 2007;388:1475–1494.
7. Romano G, Barbera N, Lombardo I. Hair testing for drugs of abuse:
evaluation of external cocaine contamination and risk of false positives.
Forensic Sci Int. 2001;123:119–129.
8. Romano G, Barbera N, Spadaro G, et al. Determination of drugs of abuse
in hair: evaluation of external heroin contamination and risk of false
positives. Forensic Sci Int. 2003;131:98–102.
9. Lucas AC, Bermejo AM, Tabernero MJ, et al. Use of solid-phase micro-
extraction (SPME) for the determination of methadone and EDDP in
human hair by GC-MS. Forensic Sci Int. 2000;107:225–232.
10. Baumgartner WA, Hill VA. Sample preparation techniques. Forensic Sci
Int. 1993;63:121–135; discussion 137–143.
11. Musshoff F, Lachenmeier K, Lichtermann D, et al. Cocaine and opiate
concentrations in hair from subjects in a heroin maintenance program in
comparison to a methadone substituted group. Int J Legal Med. 2009;
123:363–369.
12. Uhl M, Sachs H. Cannabinoids in hair: strategy to prove marijuana/
hashish consumption. Forensic Sci Int. 2004;145:143–147.
13. Kintz P, Mangin P. Simultaneous determination of opiates, cocaine and
major metabolites of cocaine in human hair by gas chromotography/mass
spectrometry (GC/MS). Forensic Sci Int. 1995;73:93–100.
14. Kronstrand R, Grundin R, Jonsson J. Incidence of opiates, amphet-
amines, and cocaine in hair and blood in fatal cases of heroin overdose.
Forensic Sci Int. 1998;92:29–38.
15. Di Corcia D, D’Urso F, Gerace E, et al. Simultaneous determination in hair
of multiclass drugs of abuse (including THC) by ultra-high performance
liquid chromatography–tandem mass spectrometry. J Chromatogr B Analyt
Technol Biomed Life Sci. 2012;899:154–159.
16. Hill V, Cairns T, Schaffer M. Hair analysis for cocaine: factors in
laboratory contamination studies and their relevance to proficiency
sample preparation and hair testing practices. Forensic Sci Int. 2008;
176:23–33.
17. Romolo FS, Rotolo MC, Palmi I, et al. Optimized conditions for simul-
taneous determination of opiates, cocaine and benzoylecgonine in hair
samples by GC–MS. Forensic Sci Int. 2003;138:17–26.
18. Barroso M, Dias M, Vieira DN, et al. Simultaneous quantitation of
morphine, 6-acetylmorphine, codeine, 6-acetylcodeine and tramadol in
hair using mixed-mode solid-phase extraction and gas chromatography-
mass spectrometry. Anal Bioanal Chem. 2010;396:3059–3069.
19. Bush DM. The U.S. Mandatory Guidelines for Federal Workplace Drug
Testing Programs: current status and future considerations. Forensic Sci
Int. 2008;174:111–119.
20. Cooper GA, Kronstrand R, Kintz P, et al. Society of Hair Testing guide-
lines for drug testing in hair. Forensic Sci Int. 2012;218:20–24.
21. United Nations Office on Drugs and Crime (UNODC). UNODC, World
Drug Report 2011. Vienna, Austria: United Nations Publication; 2011.
22. VanDerNagel JEL, Kiewik M, Jong CAJ, et al. Substance Use and
Misuse Among Intellectually Disabled Persons (SUMID). Part of the
ZonMW program: Risicogedrag en afhankelijkheid. 2008. ZonMW:
The Hague. Available at: http://www.zonmw.nl/nl/projecten/project-
detail/substance-use-and-misuse-among-intellectually-disabled-persons-sumid/
samenvatting/. Accessed on.
23. VanDerNagel J, Kiewik M, Van Dijk M, et al. Handleiding SumID-Q,
Meetinstrument voor het in kaart brengen van Middelengebruik bij men-
sen met een lichte verstandelijke beperking. SumID-Q, Substance use and
misuse in Intellectual Disability—Questionnaire. Deventer, The Nether-
lands: Tactus; 2011.
24. Society of Hair Testing. Recommendations for hair testing in forensic
cases. Forensic Sci Int. 2004;145:83–84.
25. Moore C, Coulter C, Crompton K. Determination of cocaine, benzoylec-
gonine, cocaethylene and norcocaine in human hair using solid-phase
extraction and liquid chromatography with tandem mass spectrometric
detection. J Chromatogr B Analyt Technol Biomed Life Sci. 2007;859:
208–212.
26. Lendoiro E, Quintela O, de Castro A, et al. Target screening and confir-
mation of 35 licit and illicit drugs and metabolites in hair by LC-MSMS.
Forensic Sci Int. 2012;217:207–215.
27. Bucelli F, Fratini A, Bavazzano P, et al. Quantification of drugs of abuse
and some stimulants in hair samples by liquid chromatography-electrospray
ionization ion trap mass spectrometry. J Chromatogr B Analyt Technol
Biomed Life Sci. 2009;877:3931–3936.
28. Hegstad S, Khiabani HZ, Kristoffersen L, et al. Drug screening of hair by
liquid chromatography-tandem mass spectrometry. J Anal Toxicol. 2008;
32:364–372.
29. Miller EI, Wylie FM, Oliver JS. Simultaneous detection and quantifica-
tion of amphetamines, diazepam and its metabolites, cocaine and its
metabolites, and opiates in hair by LC-ESI-MS-MS using a single extrac-
tion method. J Anal Toxicol. 2008;32:457–469.
30. Kronstrand R, Nyström I, Forsman M, et al. Hair analysis for drugs in
driver’s license regranting. A Swedish pilot study. Forensic Sci Int. 2010;
196:55–58.
Koster et al Ther Drug Monit Volume 0, Number 0, Month 2013
10 2013 Lippincott Williams & Wilkins
Copyright ªLippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
