Journal of Analytical Toxicology, Vol. 15, September/October 1991
Testing Human Hair for Drugs of Abuse. II. Identificationof Unique Cocaine Metabolites in Hair of Drug Abusers and Evaluation of Decontamination Procedures E d w a r d J. C o n e * , D a v i d Y o u s e f n e j a d , W i l l i a m D. D a r w i n , and Tom M a g u i r e
P.O. Box 5180, Addiction Research Center, National Institute on Drug Abuse, Baltimore, Maryland21224
Abstract Two unique metabolites of cocaine, cocaethylene and norcocaine, were identified by GC/MS in the hair of cocaine users. Their presence cannot be explained by environmental contamination; thus, their presence together with cocaine provides convincing evidence that cocaine is excreted in hair after active cocaine administration. The amount of cocaine in hair predominated over all metabolites generally by a factor of 5-10. Two washing procedures were evaluated for their efficiency in removal of cocaine from environmentally contaminated hair. Neither procedure completely removed cocaine, suggesting that false positives can result from environmental contamination. Analysis of the methanolic wash of the hair of cocaine users also revealed the presence of cocaine metabolites, indicating that washing removes cocaine from the interior as well as from the exterior surface of hair during decontamination procedures,
Introduction
Hair testing for drugs of abuse offers the potential for detection of drug exposure over an extended period of time. Because hair grows at an average rate of 1-1.5 cm/month ( 1), it may be possible to test hair lengths that represent months to years of potential drug exposure. With the epidemic spread of cocaine and "crack" use in the United States in the 1980s, attempts have been made to take advantage of this ostensibly hmger detection time of hair testing compared to urine testing as a means of identifying cocaine users. Early studies employing immunoassay indicated that hair testing was effective for the detection of cocaine or cocaine metabolile (benzoylecgonine) in hair segments of known users (2-4). As interest grew, more specific analytical methods were needed and assays were developed for cocaine and benzoylecgonine in hair by gas chromatography/mass spectrometry (GC/MS) (5-8) and MS/MS (8,9), Despite these early studies on hair testing for cocaine, there remains general uncertainty regarding the actual form of cocaine found in hair and its mode of entry. Most immunoassays crossreacted with cocaine and cocaine metabolites, and consequently failed to distinguish cocaine from benzoylecgonine, the major "Author to whom requests for reprints should be addressed
250
metabolite of cocaine (2-4). Also, some methods employed extraction with strong acids and bases that could convert cocaine to benzoylecgonine during sample work-up (2). In addition, the information from the early GC/MS and MS/MS studies was limited because controls that would have distinguished environmental contamination from active drug use were not included (5,7,8). Detection of a unique cocaine metabolite in hair, whose presence could not be explained by hydrolysis or environmental exposure, would unequivocally establish that internal cocaine exposure had occurred in an individual. Cocaethylene, a recently identified metabolic conversion product of cocaine and ethanol co-administration (10), would appear to meet these criteria as a unique metabolite. Cocaethylene is only formed when cocaine and ethanol are used concurrently, Because cocaethylene is not found in illicit cocaine samples ( 11), it would not be present in hair as a result of environmental contamination: thus its presence in hair could be considered a marker of cocaine exposure. Another unique metabolite which might serve as a marker of cocaine use is norcocaine, a metabolic oxidation product of cocaine (12). Studies by Cook et al. (13) have indicated that norcocaine is not formed during the pyrolysis of cocaine. The goal of the present study was the development and application of a GC/MS assay for cocaine, cocaine metabolites, and cocaethylene in hair samples obtained from active drug users. Stringent controls were included to evaluate whether wash procedures prior to assay could eliminate heavy enviromnental contamination. Ten hair samples from heavy cocaine users and 10 drug-free control hair samples were examined by the developed method. The presence of cocaethylene or norcocaine was considered to be a potential marker of cocaine exposure and evidence that cocaine excretion occurs in hair after active use. Relative amounts of cocaine, benzoylecgonine, and ecgonine methyl ester were examined to determine the actual chemical form of cocaine excreted in hair.
Materials and Methods Chemicals
Cocaine hydrochloride was obtained from Mallinckrodt, Inc. Benzoylecgonine tetrahydrate, ecgonine methyl ester HCI, ben-
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission.
Journal of Analytical Toxicology, Vol. 15, September/October 1991
zoylnorecgonine, and norcocaine HCI were obtained from the Research Technology Branch, National Institute on Drug Abuse. Cocaethylene and norcocaethylene were generous gifts from Dr. Ivy Carroll, Research Triangle Institute, Research Triangle, NC. The internal standards, 8-(methyl-2H3)-cocaine, 8-(methyl-2H3)benzoylecgonine tetrahydrate, and 8-(methyl-eH3)-ecgonine methyl ester He1 hydrate, were purchased from Sigma Chemical Co. Methanol, methylene chloride, 2-propanol, and acetonitrile (J.T. Baker, Inc.) were HPLC grade solvents. N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) with 1% trimethylchlorosilane (TMCS) was purchased from Pierce Chemical Co. All other chemicals were reagent grade. Solid-phase extraction (SPE) columns (Clean Screen | DAU, 130 mg-I mL), fritted reservoirs (4-mL), and male luer plugs were purchased from Worldwide Monitoring Corp. GC/MS autosampler microvials were purchased from Sun Brokers TM, Inc.
Hair samples Hair samples were obtained from subjects who previously had been enrolled in a 180-day outpatient maintenance and detoxification study at the Addiction Research Center. During the detoxification study, subjects received either methadone or buprenorphine. Three times per week, subjects provided urine specimens, which were tested by immunoassay (Emit | d.a.u. TM, Syva Co.) for opiates (300-ng/mL cutoff) and cocaine metabolite (300-ng/mL cutoff). After completion of the detoxification study, the subjects were requested to provide hair samples for study of their dmg usage patterns. Ten heavy cocaine users were selected for hair analysis on the basis of their high frequency of positive urines for cocaine (Table I). The interval between completion of the detoxification study and the 10 subjects' return visits to provide hair samples ranged from 4-15 months with a mean + S.E.M. of 10.2 + 1.1 months. Hair trimmings, varying in lengths from 3-10 cm, were collected from the ends of the hair shafts. Both the detoxification study and the hair study were conducted under the guidelines for the protection of human subjects, and each volunteer gave informed consent. Drug-free hair specimens were obtained from individuals who were unlikely to be cocaine users i.e., staff and children.
Hair preparation and wash procedures Hair samples (10-100 mg) were placed into preweighed 4-mL fritted reservoirs and their weights were determined. The hair in the reservoir was cut into approximately 1-mm segments with surgical scissors and the reservoir was sealed with a luer plug. Methanol (1 mL) was added and the reservoir was covered with parafilm. The reservoir was placed in a 37~ water bath for 15 min. After incubation, the reservoir was placed on a 12-port vacuum manifold (Supelco) and the methanol was collected in a glass tube. This process was repeated twice (3 x 1 mL) and the eluates were combined. Internal standard (100 ng) was added and the methanol wash solution was evaporated to dryness. The residue was extracted, derivatized, and analyzed by GC/MS assay. The decontamination procedure described by Baumgartner et al. (6) also was tested on selected control samples. Briefly, this procedure involved washing the sample in 1 mL of ethanol for 15 min at 37~ followed by three l-mL washes in phosphate buffer (pH 6) for 30 min at 37~ The ethanol and phosphate buffer washes were collected separately. Acid extraction of hair samples After the methanolic wash procedure, hair samples remained in the reservoirs. They were thoroughly dried and the reservoir tip was resealed. One mL of 0.05M sulfuric acid and internal standards (100 ng) were added, and a small stirring bar was placed in the reservoir. The top was sealed with parafilm and the reservoir was placed in a heated water bath equipped with a magnetic stirrer. The specimens were incubated overnight at 37~ with stirring. After incubation, the cooled acid extract was collected in a glass tube and saved for further extraction.
Extraction and derlvatization procedures Aliquots of the acid extract were removed and neutralized with 1.0M NaOH solution. The pH of the extract was adjusted to pH 4.0 with 1 mL of 2M sodium acetate in preparation for SPE extraction. SPE columns were mounted on an extraction vacuum manifold and conditioned with 1 mL of methanol (x2) and deionized water (x 2), respectively. Sodium acetate buffer (0.2 mL) was aspirated onto the column followed by sample. Vacuum was applied to produce a column flow of 1-2 mL/min. The columns were washed with 0.5 mL of deionized water, 1 mL of 0.1M hydrochloric acid and 2 x l mL of Table I. Characteristics and Cocaine Usage Patterns of Subjects methanol and aspirated to dryness. Analytes Participating in an Outpatient Treatment Study for Opiate Dependence* were eluted from the column with 3• 1 mL of a solution of methylene chloride-2-propanol Days (8:2) containing 2% ammonium hydroxide. Hair padicipation Cocaineuse Route 01 The eluates were evaporated to dryness under Subject Age Sex Race color in study duringstudy** administration nitrogen and reconstituted with 0.02 mL of IV acetonitrile. The samples were transferred to A 38 M B 81 180 41/77 Pos _ 0 . l - m L autosampler vials and 0.02 mL of 8 32 M 8 81 48 18/20 Pos IV BSTFA (with 1% TMCS) was added. Sample C 37 M 8 BI 29 9/9 Pos D 24 F B BI 168 71/71 Pos iN vials were sealed and heated at 60~ for 30 E 31 M B BI 66 19/26 Pos iV min. Immediately after derivatization, the F 31 M B BI 85 11/37 Pos IV samples were analyzed by GC/MS. G H I J
38 30 27 39
F F F M
B B H W
Br BI BI Br
180 31 180 170
74/76 10/10 48/77 19/58
Pos Pos Pos Pos
IN, SM IV
9 Abbreviationsare as follows: B = Black; H = Haitian;W = White; BI = Black hair; Br = Brownhair; IV = Intravenous;IN = Intranasal;SM = Smoke. '* Numberof urinestesting positivefor cocaine metabolite(EMITd.a.u. 300-ng/mLcutoff)during participation in the study.
GC/MS assay Derivatized extracts were analyzed on a Hewlett-Packard 5890A gas chromatograph coupled to a 5970B mass selective detector equipped with a 7673A automatic liquid sampler. An HP- 1 cross-linked fused-silica capil251
Journal of Analytical Toxicology, Vol. 15, September/October1991
lary column ( 12 m, 0.20-mm i.d.) with a 0.33-~m film thickness was linked to the MSD through a direct capillary interface. The multimode inlet was operated in the splitless mode with a 4-ram i.d. glass insert containing a silanized glass wool plug. The carrier gas was helium with a flow of 1 mL/min. The injector and transfer line temperatures were 250 and 280~ respectively. The oven temperature was maintained at 120~ for 1 min, then programmed to 220~ at 35~ maintained at 220 ~ C for 15 s, programmed to 250~ at 10~ maintained at 250~ for 15 s, and then programmed to 260~ at 3.5~
soaking, the samples were blotted dry, air dried, and stored in individual plastic bags and frozen until analysis.
Results Identification and quantitation of cocaine and metabolites in cocaine-users' hair Ten hair samples obtained from cocaine users (Table II) were washed with methanol, extracted in sulfuric acid solution, and analyzed for cocaine and cocaine metabolites by GC/MS. The presence of cocaine, benzoylecgonine, ecgonine methyl ester, norcocaine, and cocaethylene was established by comparison of the mass spectral patterns with authentic standards. An example of the mass spectral identification of cocaethylene in the acid extract of Subject B compared to standard cocaethylene is shown in Figure 1. A small amount of coeluting impurity (m/z 73, l l7, and 296) is present in Subject B's spectrum (Figure 1B). An additional metabolite, norcocaethylene, was detected by SIM analysis in most samples that contained cocaethylene. However, the abundance was not sufficient for full scan identification. An example of the SIM tracings for cocaethylene, benzoylecgonine, norcocaine, and norcocaethylene is shown in Figure 2. Responses are evident for the respective nor-metabolites of cocaine and cocaethylene at m/z 140 and 360 in the extract fraction (Panel 2B) and in the methanolic wash fraction (Panel 2D) from Subject B. The ion for benzoylecgonine, m& 240, also serves as a confirming ion for norcocaine (Panel 2A). Subject A's hair did not contain cocaethylene and consequently shows a lack of response at m/z 196 for cocaethylene and at m/z 360 for norcocaethylene. The quantitative analysis of cocaine and cocaine metabolites in hair was performed by first washing the samples with methanol, then extraction in sulfuric acid solution, further extraction, and GC/MS assay. All samples, calibrators, and controls were processed in the same manner. The assay had a limit of detection of approximately 0.1 ng/mg with a 50-mg hair sample for all analytes. Recovery of cocaine that was added to control hair (after the wash step) was approximately 90% with a 10% conversion to benzoylecgonine. The efficiency of cocaine recover~
The MSD was operated in the selected ion monitoring (SIM) mode. The following ions were monitored for each compound at their respective retention times (Rt): ecgonine methyl ester, m/z 82, (96), 271, Rt = 3.50 min; [2H3]-ecgonine methyl ester, m/z 99, Rt = 3.49 min; cocaine, m/z 82, (182), 303, Rt = 6.36 min; [2H3]-cocaine, m/z 85, (185), R t = 6.35 min; cocaethylene, m/z 82, (196), Rt = 6.71 min; benzoylecgonine, m/z 82, (240), 361, Rt = 6.85 min ;[2H3] BE, m/z (243), Rt = 6.83 min; norcocaine, m/z (140), 240, 361, R t = 6.99 min; norcocaethylene, 140 (360), 361, R t = 7.29 min; benzoylnorecgonine, 140 (404), Rt = 7.59 min. Ions used for quantitation are shown in parenthesis. Analytes were identified based on comparison of retention times and relative abundance of two confirming ions to the corresponding values of authentic standards run on a daily basis. A dwell time of 10 ms was used for ion acquisition. Standard curves (12.5-1,000 ng) were constructed based on ion peakheight ratios of analyte to their respective deuterated analogs. Preparation of environmentally exposed cocaine hair samples Hair samples exposed to cocaine vapor (simulated "crack" exposure) were prepared by suspending bundles of control hair (approximately 100 rag) inside a plexi-glass box (2x3x3 It.) in which cocaine base (10 rag) was heated at 200~ The hair bundles were suspended at a distance of 10 in. above the heated cocaine tbr 1 h. Similar hair samples suspended outside the box were used as drug-free controls. After exposure, the samples were stored in individual plaslic bags and frozen until analysis. Hair samples exposed to aqueous solutions of cocaine hydrochloride were prepared by soaking bundles of control hair (approximately 100 mg) in solutions of cocaine hydrochloride ( 1 mg/mL, tap water) at room temperature for 48 h. Similar hair samples were prepared by soaking in drug-free tap water. After
Table II. C G / M S Assay of Wash and Acid Extract of Cocaine Users' Hair* Cocaine (ng/mg) Subject
Wash
A B C D E F G H 1 J
23.4 8.2 40.7 12.0 36.3 31.1 178.t 63.3 59.0 10,0 46.2 15.9
Mean SEM
Extract
Benzoylecgonine Ecgoninemethyl ester (ng/mg) (ng/mg)
Cocaethylene (ng/mg)
Wash
Extract
Wash
Extract
Wash
Extract
Wash
Extract
6.8 9.3 9.7 19,2 7.8 14,5 15.8 11,4 6.6 6.4
2.0 0.4 3.5 1.2 2.4 2.7 24,9 7.7 1.6 0.5
0.6 0.9 1.2 1.3 0.8 1.5 2.5 2.2 0.3 0.9
T T 1.9 T 0,3 0.2 2.9 1,0 0,2 T
0.6 1.9 1.3 1.9 0.5 1.8 0 T O T
T T T T 0.9 T 2.6 0.8 T 0
T 0.5 T T 0,5 0.7 T 0.5 T T
0 1.1 11.9 0 3.8 1.4 0 2.5 T T
0 1.3 2.6 T 1.2 1.0 0 0.6 T 0,4
10.8 1.4
4.7 2.3
1,2 0.2
0.7 0.3
0.8 0.3
0.4 0.3
0.2 0.1
2.1 1.2
0.7 0.3
9 Some analytes were detected in trace (T) amounts below the GC/MS calibration curve.
252
Norcocaine (ng/mg)
Norcocaethylene (rig/rag) Wash
Extract
Journal of Analytical Toxicology, Vol. 15, September/October 1991
s2
lOO 1
A) Cocaethylene Standard 196
5O
iL,r
77
o .t, 82
loo
B)
I
50
LL ...... ,
,
',
-
Subject B, Hair Extract, Rt = 6.7 min
196
I Ira
Analysis of environmentally exposed cocaine hair samples
o ~,llililii,lii.i,.,.,!.,iLllL,,,,iio.,,, 100
3117
27~,
lrJLl:ii
60
L
272 212
,-
194
Im I
L
317
140
180
..........., 220
260
, J ~......, 300
340
M/Z Figure 1. Mass spectra of cocaethylene standard and cocaethylene extracted from a cocaine-user's hair sample. Im= coeluting impurity in the spectrum of Subject B's hair sample.
A) STDS
CI 111 A 9 IIm
B) Sub. B Extract
I! i ~ I~ /
CI ill ~ m II m
[
~ m
i
'
C) Sub. A
Extract
I 1 ~
i'
m
m
from cocaine user's hair was not determined, although re-extraction indicated that recovery was incomplete. Quantitative results from the analysis of the wash fraction and the acid extract fraction from 10 cocaine users are summarized in Table II, The amount of cocaine and metabolites found in the wash fraction was generally equal to or higher than that found in the acid extract. Amounts of cocaine in the acid extract ranged from 6.4 to 19.2 ng/mg with a mean + SEM of 10.8 + 1.4 ng/mg. Benzoylecgonine ranged from 0.3 to 2.5 ng/mg with a mean + SEM of 1.2 + 0.2 ng/mg. The ratio of cocaine to benzoylecgonine in the acid extract (uncorrected for chemical conversion of cocaine to benzoylecgonine) ranged from 5.2 to 22.0 with a mean + SEM of 10.5 + 1.5, Ecgonine methyl ester ranged from 0.6 to 1.9 ng/mg. Norcocaine was assayed in 4 samples in amounts ranging from 0.5 to 0.7 ng/mg. Cocaethylene was measured in 6 samples in amounts ranging from 0.4 to 2.6 ng/mg. Traces of norcocaethylene were detected in those specimens containing cocaethylene. Ten drug-free control samples were processed in the same manner as the cocaine users' samples. The acid extracts from these samples were negative for cocaine and metabolites,
0 m
I
C~ III D) Sub. B / mW
Hair samples (ca. 100 mg) exposed to cocaine vapor were prewashed by two methods in an attempt to remove surface contamination. The amount of cocaine remaining on hair was measured by SIM at m/z 182; D3-cocaine was monitored at m/z 185 (Figure 3). Standard cocaine appeared at a retention time of 6.28 rain. (Figure 3A). As shown in Figure 3B, control hair exposed to drug-free room air was negative for cocaine. Samples exposed to cocaine vapor remained heavily contaminated with cocaine after washing with methanol (Figure 3C). The total amount of cocaine remaining on the hair samples ranged from 510-3,430 ng by the methanol procedure and 1,851-9,485 ng by the Baumgartner method (Table III). Washing with methanol ( 1 mL x 3) was effective in removing more than 90% of the cocaine from hair, whereas use of the alternate procedure reported by Baumgartner et al. (6) removed an average of 81%. Hair samples soaked in 1 mg/mL of aqueous cocaine hydrochloride remained grossly contaminated with cocaine after both wash procedures. No attempt was made to quantitate the amount of cocaine remaining in hair because all GC/MS responses were saturated (Figure 3D).
WASH
Discussion M/ZI~ --
It
Minutes
Figure 2. SIM recordings of standards, hair extracts, and methanolic wash for cocaethylene (CE), benzoylecgonine (BE), norcocaine (NC), and norcocaethylene (NCE). Panel A illustrates the response for a control hair extract containing 25 ng (total) of BE and 12.5 ng (total) of CE, NC, and NCE. Panels B and C illustrate the response for acid extracts of hair samples from Subjects B and A, respectively. Panel D represents the response for the methanolic wash prior to acid extraction of Subject B's hair sample.
Hair analysis for cocaine has been stated to be an efficient means of providing "... useful information on cocaine consumption by cocaine addicts ..." (2). Accordingly, Baumgartner reported finding a linear response between cocaine metabolite (presumably benzoylecgonine) in mouse hair and the administered dose (6). In the same article, he also reported finding a "reasonably good correlation" between cocaine metabolite in the hair of patients, probationers, and parolees and their self-reported amounts of cocaine use (6). In contrast, Kidwell (9) reported finding a large excess of intact cocaine over benzoylecgonine and ecgonine in the hair of users by tandem MS analysis under pyrolysis conditions. In the present study, we sought to clarify the nature of the chemical species that is excreted in hair
253
Journal of Analytical Toxicology, Vol. 15, September/October1991
same time or within a few hours (14). Hence, the detection of cocaethylene in 8 of 10 heavy cocaine users' hair is not surprising. The detection of traces of norcocaethylene, the Ndemethylated metabolite of cocaethylene, in the hair samples containing cocaethylene further supports the current finding of cocaethylene. The washing procedure was included in an attempt to eliminate potential surface contamination from environmental cocaine. Human head hair could be exposed to a variety of sources of cocaine free base and cocaine hydrochloride from the environment. Smoking "crack" involves the vaporization of base cocaine. Small amounts of "sidestream" crack vapor can be released into the atmosphere. This allows the possibility of adsorption of cocaine onto an individual's hair located nearby. Certainly, the smoker's hair would be contaminated in the same process. Alternately, cocaine hydrochloride dust particles in the air and on articles of clothing and money would present the opportunity for hair contamination from the powder and aqueous solutions. Although the environmentally contaminated controls utilized in this study simulated examples of extreme exposures from cocaine base and aqueous solutions of cocaine hydrochloride, other investigators have employed even more exaggerated conditions. Baumgartner and Hill (15) produced contaminated hair specimens for study by suspending locks of hair in a 1-L glass flask filled with vapor from 500 mg of cocaine freebase.
after cocaine use. This required the development of a sensitive and specific GC/MS assay for cocaine and metabolites for use with small amounts of hair samples. Controls were included to evaluate the potential role of environmental contamination of cocaine in hair analysis. Analysis of hair samples from the 10 documented cocaine users provided evidence for the presence of cocaine in each subject's hair, whereas negative findings were obtained for the 10 drug-free controls. The mean ratio of cocaine to benzoylecgonine was 10.5. Only traces of ecgonine methyl ester were detected. These findings agree with Kidwell's findings (9) that cocaine is the principal analyte in hair and occurs in much higher abundance than other cocaine metabolites. The detection of norcocaine and cocaethylene along with cocaine by GC/MS in the present study provided convincing proof that cocaine is excreted in hair and rules out the possibility of environmental contamination as an alternate explanation for cocaine's occurrence. Neither of these compounds occur in appreciable amounts in illicit cocaine and their presence can only be explained by human biotransformation of cocaine to these unique metabolites. Cocaine is oxidatively metabolized to norcocaine (12) and cocaethylene has been shown to arise as a result of the concurrent use of cocaine and alcohol (10). It is known that alcohol is often abused along with cocaine. A recent report indicates that approximately 85% of cocaine users polled in the 1985 National Survey on Drug Abuse used ethanol at the B) Control Vapor
A) Cocaine Std.
D) Aqueous Cocaine
C) Cocaine Vapor
;\
m/z 182
-
-
/
~-
m/z 185 6.'10
t 6.20
6.30
i 6.40
6.'10
6.20
6"30
m/z 182 m/z 185
6.40
61o
6.'~o 6.3o 6.40
,',0
6.'~0 6.30 ,.40
Minutes posed to room air and cocaine vapor (10 mg in a 2 x 3 x 3 ft box), respectively. Panel D illustrates the response for the acid extract (after methanol wash) for cocaine-free control hair soaked for 48 h in aqueous cocaine solution (1 mg/mL). Note the saturated response for cocaine in Panel D.
Figure 3. SIM recordings of a standard cocaine hair extract and environmentally contaminated hair samples. Panel A represents the response for standard cocaine (50 ng) at m/z182 (Rt = 6.29 rain) and deuterated cocaine (100 ng) at m/z185 (Rt = 6.28 min). Panels B and C illustrate the response for the acid extract (after methanol wash) for cocaine-free control hair ex-
Table III. Efficiency of Wash Procedures for Removal of Cocaine from Control Hair Specimens That Were Contaminated with Cocaine Vapor* Total CocaineRecovered(ng)
Percent cocaineremovedby washing
Wash fraction Wash Method Methanol
Ethanol-phosphate buffer (6)
Alcohol
Buffer
Acid extract fraction
Alcoholwash
Buffer wash
Total
11,052 30,954 118,320
-
510 1,019 3,430
95.6 96.8 97.2
-
95.6 96.8 97.2
9,238 31,160 t7,208
1,057 7,396 4,978
1,851 9,485 6,288
76.1 64.9 60.4
8.7 15.4 17.5
84.8 80.3 77.9
9 Triplicate drug-free hair specimens (Caucasoid, brown, ca. 100 mg each) were analyzed by two wash procedures as described in Methods section. The data indicate the total amount of cocaine recovered in ng by each procedure.
254
Journal of Analytical Toxicology, Vol. 15, September/October 1991
The methanol washing procedure used in the present study was more than 90% effective in removing vaporized cocaine, but sufficient cocaine residue remained to produce false positive results. After soaking in an aqueous solution of cocaine, the methanol wash procedure was totally inadequate in removing cocaine; hair samples remained extremely contaminated. The washing procedure reported by Baumgartner et al. (6) fared no better in removing cocaine contamination from these samples. Additional wash-out kinetic criteria have been proposed by Baumgartner and Hill (15) as a means of distinguishing external cocaine contamination from active drug use; however, these criteria remain untested by other investigators. Other studies have included a variety of conditions in attempts to remove drug contamination. Marigo et al. (16) reported that serial washings with ethyl ether and dilute HC1 successfully removed "loosely bound morphine." In contrast, Kidwell (9) reported the failure of a variety of wash solutions in removing phencyclidine from contaminated hair. Analysis o f the methanolic wash for cocaine metabolites produced evidence that the wash procedure removed more than just external cocaine contamination. The presence of norcocaine and cocaethylene along with cocaine and benzoylecgonine in amounts generally exceeding those found in the extract indicated that interior drug components were removed in the wash. Clearly in the present procedure, the "wash" procedure could more aptly be called an extraction. The depth of penetration by the "wash" solution is likely to be dependent upon the condition of the hair undergoing testing. Highly damaged hair is likely to be more porous and yield greater amount of interior drug during washing. If this is true, the amount of drug remaining in the extract is likely to be distorted and bear little relationship to dosing patterns. Based on the present data, it seems unlikely that there is a clear demarcation in hair between environmental drug exposure and drug residues resulting from active use. The elimination of surface contamination from hair prior to its analysis for drugs of abuse is important, given the possibilities of environmental contamination. Unfortunately, in the present effort, the techniques employed were not effective. It may be necessary to adopt an alternate approach in the development of hair analysis technology by requiring the identification of a unique metabolite whose presence can only be explained by internal drug exposure with subsequent biotransformation. The presence of cocaine, benzoylecgonine, and ecgonine methyl ester could potentially be explained by environmental contamination with subsequent hydrolysis. In sharp contrast, the presence of either norcocaine or cocaethylene in an individual's hair would appear to constitute sufficient evidence of internal cocaine exposure.
References 1. M. Saitoh, M. Uzuka, and M. Sakamoto. Rate of hair growth. In Advances in Biology of Skin. Vo/. IX Hair Growth. W. Montagna and R.L Dobson, Eds., Pergamon, Oxford, pp. 183-201 (1969). 2. D. Valente, M. Cassini, M. Pigliapochi, and G. Vansetti. Hair as the sample in assessing morphine and cocaine addiction. Olin. Chem. 27:1952-53 (1981). 3. W.A. Baumgartner, C.T. Black, P.F. Jones, and W.H. Blahd. Radioimmunoassay of cocaine in hair: Concise communication. J. NucL Med. 23." 790-92 (1982). 4. F.P. Smith and R.H. Liu. Detection of cocaine metabolite in perspiration stain, menstrual bloodstain, and hair. J. Forens. Sci. 31." 1269-73 (1986). 5. S.A. Reuschel and F.P. Smith. Benzoylecgonine (cocaine metabolite) detection in hair samples of jail detainees using RIA and GC/MS. J. Forens. Sci., in press, 1991. 6. W.A. Baumgartner, V.A. Hill, and W.D. Blahd. Hair analysis for drugs of abuse. J. Forens. Sci. 34:1433-53 (1989). 7. S. Balabanova and J. Homoki. Determination of cocaine in human hair by gas chromatography/mass spectrometry. Z. Rechtsmed. 98:235-40 (1987). 8. RM. Martz. The identification of cocaine in hair by GC/MS and MS/MS. Crime Lab. Digest 15:67-73 (1988). 9. D.A. Kidwell. Analysis of drugs of abuse in hair by tandem mass spectrometry. In Proceedings, 36th American Society of Mass Spectrometry Conference on Mass Spectrometry and Allied Topics, San Francisco, CA, 1989, pp. 1364-65. 10. W.L Hearn, D.D. Flynn, G.W. Hime, S. Rose, J.C. Cofino, E. Mantero-Atienza, C.V. Wetli, and D.C. Mash. Cocaethylene: A unique metabolite displays high affinity for the dopamine transporter. J. Neurochem. 56:698-701 (1991). 11. Personal communication, W. Lee Hearn. 12. T. Inaba, D.J. Stewart, and W. Kalow. Metabolism of cocaine in man. Clin. Pharmacol. Ther. 23:547-52 (1978). 13. C.E. Cook, A.R. Jeffcoat, and M. Perez-Reyes. Pharmacokinetics studies of cocaine and phencyclidine in man. In Pharmacokinetics and Pharmacodynamics of Psychoactive Drugs, G. Barnett and C.N. Chiang, Eds., Biomedical Publications, Foster City, CA, 1985, pp. 48-74. 14. B.F. Grant and T.C. Harford. Concurrent and simultaneous use of alcohol with cocaine: Results of national survey. Drug Alcohol Dep. 25:97-104 (1990). 15. W.A. Baumgartner and V.A. Hill. Hair analysis for drugs of abuse: Decontamination issues. In Proceedings of the 2rid International Congress of Therapeutic Drug Monitoring and Toxicology, Oct. 9-12, 1990, Barcelona, Spain, in press. 16. M Marigo, F. Tagliaro, C. Poiesi, S. Lafisca, and C. Neri. Determination of morphine in the hair of heroin addicts by high performance liquid chromatography with fluorimetric detection. J. Anal. ToxicoL 10:158-61 (1986).
255
Testing Human Hair for Drugs of Abuse. II. Identificationof Unique Cocaine Metabolites in Hair of Drug Abusers and Evaluation of Decontamination Procedures E d w a r d J. C o n e * , D a v i d Y o u s e f n e j a d , W i l l i a m D. D a r w i n , and Tom M a g u i r e
P.O. Box 5180, Addiction Research Center, National Institute on Drug Abuse, Baltimore, Maryland21224
Abstract Two unique metabolites of cocaine, cocaethylene and norcocaine, were identified by GC/MS in the hair of cocaine users. Their presence cannot be explained by environmental contamination; thus, their presence together with cocaine provides convincing evidence that cocaine is excreted in hair after active cocaine administration. The amount of cocaine in hair predominated over all metabolites generally by a factor of 5-10. Two washing procedures were evaluated for their efficiency in removal of cocaine from environmentally contaminated hair. Neither procedure completely removed cocaine, suggesting that false positives can result from environmental contamination. Analysis of the methanolic wash of the hair of cocaine users also revealed the presence of cocaine metabolites, indicating that washing removes cocaine from the interior as well as from the exterior surface of hair during decontamination procedures,
Introduction
Hair testing for drugs of abuse offers the potential for detection of drug exposure over an extended period of time. Because hair grows at an average rate of 1-1.5 cm/month ( 1), it may be possible to test hair lengths that represent months to years of potential drug exposure. With the epidemic spread of cocaine and "crack" use in the United States in the 1980s, attempts have been made to take advantage of this ostensibly hmger detection time of hair testing compared to urine testing as a means of identifying cocaine users. Early studies employing immunoassay indicated that hair testing was effective for the detection of cocaine or cocaine metabolile (benzoylecgonine) in hair segments of known users (2-4). As interest grew, more specific analytical methods were needed and assays were developed for cocaine and benzoylecgonine in hair by gas chromatography/mass spectrometry (GC/MS) (5-8) and MS/MS (8,9), Despite these early studies on hair testing for cocaine, there remains general uncertainty regarding the actual form of cocaine found in hair and its mode of entry. Most immunoassays crossreacted with cocaine and cocaine metabolites, and consequently failed to distinguish cocaine from benzoylecgonine, the major "Author to whom requests for reprints should be addressed
250
metabolite of cocaine (2-4). Also, some methods employed extraction with strong acids and bases that could convert cocaine to benzoylecgonine during sample work-up (2). In addition, the information from the early GC/MS and MS/MS studies was limited because controls that would have distinguished environmental contamination from active drug use were not included (5,7,8). Detection of a unique cocaine metabolite in hair, whose presence could not be explained by hydrolysis or environmental exposure, would unequivocally establish that internal cocaine exposure had occurred in an individual. Cocaethylene, a recently identified metabolic conversion product of cocaine and ethanol co-administration (10), would appear to meet these criteria as a unique metabolite. Cocaethylene is only formed when cocaine and ethanol are used concurrently, Because cocaethylene is not found in illicit cocaine samples ( 11), it would not be present in hair as a result of environmental contamination: thus its presence in hair could be considered a marker of cocaine exposure. Another unique metabolite which might serve as a marker of cocaine use is norcocaine, a metabolic oxidation product of cocaine (12). Studies by Cook et al. (13) have indicated that norcocaine is not formed during the pyrolysis of cocaine. The goal of the present study was the development and application of a GC/MS assay for cocaine, cocaine metabolites, and cocaethylene in hair samples obtained from active drug users. Stringent controls were included to evaluate whether wash procedures prior to assay could eliminate heavy enviromnental contamination. Ten hair samples from heavy cocaine users and 10 drug-free control hair samples were examined by the developed method. The presence of cocaethylene or norcocaine was considered to be a potential marker of cocaine exposure and evidence that cocaine excretion occurs in hair after active use. Relative amounts of cocaine, benzoylecgonine, and ecgonine methyl ester were examined to determine the actual chemical form of cocaine excreted in hair.
Materials and Methods Chemicals
Cocaine hydrochloride was obtained from Mallinckrodt, Inc. Benzoylecgonine tetrahydrate, ecgonine methyl ester HCI, ben-
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission.
Journal of Analytical Toxicology, Vol. 15, September/October 1991
zoylnorecgonine, and norcocaine HCI were obtained from the Research Technology Branch, National Institute on Drug Abuse. Cocaethylene and norcocaethylene were generous gifts from Dr. Ivy Carroll, Research Triangle Institute, Research Triangle, NC. The internal standards, 8-(methyl-2H3)-cocaine, 8-(methyl-2H3)benzoylecgonine tetrahydrate, and 8-(methyl-eH3)-ecgonine methyl ester He1 hydrate, were purchased from Sigma Chemical Co. Methanol, methylene chloride, 2-propanol, and acetonitrile (J.T. Baker, Inc.) were HPLC grade solvents. N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) with 1% trimethylchlorosilane (TMCS) was purchased from Pierce Chemical Co. All other chemicals were reagent grade. Solid-phase extraction (SPE) columns (Clean Screen | DAU, 130 mg-I mL), fritted reservoirs (4-mL), and male luer plugs were purchased from Worldwide Monitoring Corp. GC/MS autosampler microvials were purchased from Sun Brokers TM, Inc.
Hair samples Hair samples were obtained from subjects who previously had been enrolled in a 180-day outpatient maintenance and detoxification study at the Addiction Research Center. During the detoxification study, subjects received either methadone or buprenorphine. Three times per week, subjects provided urine specimens, which were tested by immunoassay (Emit | d.a.u. TM, Syva Co.) for opiates (300-ng/mL cutoff) and cocaine metabolite (300-ng/mL cutoff). After completion of the detoxification study, the subjects were requested to provide hair samples for study of their dmg usage patterns. Ten heavy cocaine users were selected for hair analysis on the basis of their high frequency of positive urines for cocaine (Table I). The interval between completion of the detoxification study and the 10 subjects' return visits to provide hair samples ranged from 4-15 months with a mean + S.E.M. of 10.2 + 1.1 months. Hair trimmings, varying in lengths from 3-10 cm, were collected from the ends of the hair shafts. Both the detoxification study and the hair study were conducted under the guidelines for the protection of human subjects, and each volunteer gave informed consent. Drug-free hair specimens were obtained from individuals who were unlikely to be cocaine users i.e., staff and children.
Hair preparation and wash procedures Hair samples (10-100 mg) were placed into preweighed 4-mL fritted reservoirs and their weights were determined. The hair in the reservoir was cut into approximately 1-mm segments with surgical scissors and the reservoir was sealed with a luer plug. Methanol (1 mL) was added and the reservoir was covered with parafilm. The reservoir was placed in a 37~ water bath for 15 min. After incubation, the reservoir was placed on a 12-port vacuum manifold (Supelco) and the methanol was collected in a glass tube. This process was repeated twice (3 x 1 mL) and the eluates were combined. Internal standard (100 ng) was added and the methanol wash solution was evaporated to dryness. The residue was extracted, derivatized, and analyzed by GC/MS assay. The decontamination procedure described by Baumgartner et al. (6) also was tested on selected control samples. Briefly, this procedure involved washing the sample in 1 mL of ethanol for 15 min at 37~ followed by three l-mL washes in phosphate buffer (pH 6) for 30 min at 37~ The ethanol and phosphate buffer washes were collected separately. Acid extraction of hair samples After the methanolic wash procedure, hair samples remained in the reservoirs. They were thoroughly dried and the reservoir tip was resealed. One mL of 0.05M sulfuric acid and internal standards (100 ng) were added, and a small stirring bar was placed in the reservoir. The top was sealed with parafilm and the reservoir was placed in a heated water bath equipped with a magnetic stirrer. The specimens were incubated overnight at 37~ with stirring. After incubation, the cooled acid extract was collected in a glass tube and saved for further extraction.
Extraction and derlvatization procedures Aliquots of the acid extract were removed and neutralized with 1.0M NaOH solution. The pH of the extract was adjusted to pH 4.0 with 1 mL of 2M sodium acetate in preparation for SPE extraction. SPE columns were mounted on an extraction vacuum manifold and conditioned with 1 mL of methanol (x2) and deionized water (x 2), respectively. Sodium acetate buffer (0.2 mL) was aspirated onto the column followed by sample. Vacuum was applied to produce a column flow of 1-2 mL/min. The columns were washed with 0.5 mL of deionized water, 1 mL of 0.1M hydrochloric acid and 2 x l mL of Table I. Characteristics and Cocaine Usage Patterns of Subjects methanol and aspirated to dryness. Analytes Participating in an Outpatient Treatment Study for Opiate Dependence* were eluted from the column with 3• 1 mL of a solution of methylene chloride-2-propanol Days (8:2) containing 2% ammonium hydroxide. Hair padicipation Cocaineuse Route 01 The eluates were evaporated to dryness under Subject Age Sex Race color in study duringstudy** administration nitrogen and reconstituted with 0.02 mL of IV acetonitrile. The samples were transferred to A 38 M B 81 180 41/77 Pos _ 0 . l - m L autosampler vials and 0.02 mL of 8 32 M 8 81 48 18/20 Pos IV BSTFA (with 1% TMCS) was added. Sample C 37 M 8 BI 29 9/9 Pos D 24 F B BI 168 71/71 Pos iN vials were sealed and heated at 60~ for 30 E 31 M B BI 66 19/26 Pos iV min. Immediately after derivatization, the F 31 M B BI 85 11/37 Pos IV samples were analyzed by GC/MS. G H I J
38 30 27 39
F F F M
B B H W
Br BI BI Br
180 31 180 170
74/76 10/10 48/77 19/58
Pos Pos Pos Pos
IN, SM IV
9 Abbreviationsare as follows: B = Black; H = Haitian;W = White; BI = Black hair; Br = Brownhair; IV = Intravenous;IN = Intranasal;SM = Smoke. '* Numberof urinestesting positivefor cocaine metabolite(EMITd.a.u. 300-ng/mLcutoff)during participation in the study.
GC/MS assay Derivatized extracts were analyzed on a Hewlett-Packard 5890A gas chromatograph coupled to a 5970B mass selective detector equipped with a 7673A automatic liquid sampler. An HP- 1 cross-linked fused-silica capil251
Journal of Analytical Toxicology, Vol. 15, September/October1991
lary column ( 12 m, 0.20-mm i.d.) with a 0.33-~m film thickness was linked to the MSD through a direct capillary interface. The multimode inlet was operated in the splitless mode with a 4-ram i.d. glass insert containing a silanized glass wool plug. The carrier gas was helium with a flow of 1 mL/min. The injector and transfer line temperatures were 250 and 280~ respectively. The oven temperature was maintained at 120~ for 1 min, then programmed to 220~ at 35~ maintained at 220 ~ C for 15 s, programmed to 250~ at 10~ maintained at 250~ for 15 s, and then programmed to 260~ at 3.5~
soaking, the samples were blotted dry, air dried, and stored in individual plastic bags and frozen until analysis.
Results Identification and quantitation of cocaine and metabolites in cocaine-users' hair Ten hair samples obtained from cocaine users (Table II) were washed with methanol, extracted in sulfuric acid solution, and analyzed for cocaine and cocaine metabolites by GC/MS. The presence of cocaine, benzoylecgonine, ecgonine methyl ester, norcocaine, and cocaethylene was established by comparison of the mass spectral patterns with authentic standards. An example of the mass spectral identification of cocaethylene in the acid extract of Subject B compared to standard cocaethylene is shown in Figure 1. A small amount of coeluting impurity (m/z 73, l l7, and 296) is present in Subject B's spectrum (Figure 1B). An additional metabolite, norcocaethylene, was detected by SIM analysis in most samples that contained cocaethylene. However, the abundance was not sufficient for full scan identification. An example of the SIM tracings for cocaethylene, benzoylecgonine, norcocaine, and norcocaethylene is shown in Figure 2. Responses are evident for the respective nor-metabolites of cocaine and cocaethylene at m/z 140 and 360 in the extract fraction (Panel 2B) and in the methanolic wash fraction (Panel 2D) from Subject B. The ion for benzoylecgonine, m& 240, also serves as a confirming ion for norcocaine (Panel 2A). Subject A's hair did not contain cocaethylene and consequently shows a lack of response at m/z 196 for cocaethylene and at m/z 360 for norcocaethylene. The quantitative analysis of cocaine and cocaine metabolites in hair was performed by first washing the samples with methanol, then extraction in sulfuric acid solution, further extraction, and GC/MS assay. All samples, calibrators, and controls were processed in the same manner. The assay had a limit of detection of approximately 0.1 ng/mg with a 50-mg hair sample for all analytes. Recovery of cocaine that was added to control hair (after the wash step) was approximately 90% with a 10% conversion to benzoylecgonine. The efficiency of cocaine recover~
The MSD was operated in the selected ion monitoring (SIM) mode. The following ions were monitored for each compound at their respective retention times (Rt): ecgonine methyl ester, m/z 82, (96), 271, Rt = 3.50 min; [2H3]-ecgonine methyl ester, m/z 99, Rt = 3.49 min; cocaine, m/z 82, (182), 303, Rt = 6.36 min; [2H3]-cocaine, m/z 85, (185), R t = 6.35 min; cocaethylene, m/z 82, (196), Rt = 6.71 min; benzoylecgonine, m/z 82, (240), 361, Rt = 6.85 min ;[2H3] BE, m/z (243), Rt = 6.83 min; norcocaine, m/z (140), 240, 361, R t = 6.99 min; norcocaethylene, 140 (360), 361, R t = 7.29 min; benzoylnorecgonine, 140 (404), Rt = 7.59 min. Ions used for quantitation are shown in parenthesis. Analytes were identified based on comparison of retention times and relative abundance of two confirming ions to the corresponding values of authentic standards run on a daily basis. A dwell time of 10 ms was used for ion acquisition. Standard curves (12.5-1,000 ng) were constructed based on ion peakheight ratios of analyte to their respective deuterated analogs. Preparation of environmentally exposed cocaine hair samples Hair samples exposed to cocaine vapor (simulated "crack" exposure) were prepared by suspending bundles of control hair (approximately 100 rag) inside a plexi-glass box (2x3x3 It.) in which cocaine base (10 rag) was heated at 200~ The hair bundles were suspended at a distance of 10 in. above the heated cocaine tbr 1 h. Similar hair samples suspended outside the box were used as drug-free controls. After exposure, the samples were stored in individual plaslic bags and frozen until analysis. Hair samples exposed to aqueous solutions of cocaine hydrochloride were prepared by soaking bundles of control hair (approximately 100 mg) in solutions of cocaine hydrochloride ( 1 mg/mL, tap water) at room temperature for 48 h. Similar hair samples were prepared by soaking in drug-free tap water. After
Table II. C G / M S Assay of Wash and Acid Extract of Cocaine Users' Hair* Cocaine (ng/mg) Subject
Wash
A B C D E F G H 1 J
23.4 8.2 40.7 12.0 36.3 31.1 178.t 63.3 59.0 10,0 46.2 15.9
Mean SEM
Extract
Benzoylecgonine Ecgoninemethyl ester (ng/mg) (ng/mg)
Cocaethylene (ng/mg)
Wash
Extract
Wash
Extract
Wash
Extract
Wash
Extract
6.8 9.3 9.7 19,2 7.8 14,5 15.8 11,4 6.6 6.4
2.0 0.4 3.5 1.2 2.4 2.7 24,9 7.7 1.6 0.5
0.6 0.9 1.2 1.3 0.8 1.5 2.5 2.2 0.3 0.9
T T 1.9 T 0,3 0.2 2.9 1,0 0,2 T
0.6 1.9 1.3 1.9 0.5 1.8 0 T O T
T T T T 0.9 T 2.6 0.8 T 0
T 0.5 T T 0,5 0.7 T 0.5 T T
0 1.1 11.9 0 3.8 1.4 0 2.5 T T
0 1.3 2.6 T 1.2 1.0 0 0.6 T 0,4
10.8 1.4
4.7 2.3
1,2 0.2
0.7 0.3
0.8 0.3
0.4 0.3
0.2 0.1
2.1 1.2
0.7 0.3
9 Some analytes were detected in trace (T) amounts below the GC/MS calibration curve.
252
Norcocaine (ng/mg)
Norcocaethylene (rig/rag) Wash
Extract
Journal of Analytical Toxicology, Vol. 15, September/October 1991
s2
lOO 1
A) Cocaethylene Standard 196
5O
iL,r
77
o .t, 82
loo
B)
I
50
LL ...... ,
,
',
-
Subject B, Hair Extract, Rt = 6.7 min
196
I Ira
Analysis of environmentally exposed cocaine hair samples
o ~,llililii,lii.i,.,.,!.,iLllL,,,,iio.,,, 100
3117
27~,
lrJLl:ii
60
L
272 212
,-
194
Im I
L
317
140
180
..........., 220
260
, J ~......, 300
340
M/Z Figure 1. Mass spectra of cocaethylene standard and cocaethylene extracted from a cocaine-user's hair sample. Im= coeluting impurity in the spectrum of Subject B's hair sample.
A) STDS
CI 111 A 9 IIm
B) Sub. B Extract
I! i ~ I~ /
CI ill ~ m II m
[
~ m
i
'
C) Sub. A
Extract
I 1 ~
i'
m
m
from cocaine user's hair was not determined, although re-extraction indicated that recovery was incomplete. Quantitative results from the analysis of the wash fraction and the acid extract fraction from 10 cocaine users are summarized in Table II, The amount of cocaine and metabolites found in the wash fraction was generally equal to or higher than that found in the acid extract. Amounts of cocaine in the acid extract ranged from 6.4 to 19.2 ng/mg with a mean + SEM of 10.8 + 1.4 ng/mg. Benzoylecgonine ranged from 0.3 to 2.5 ng/mg with a mean + SEM of 1.2 + 0.2 ng/mg. The ratio of cocaine to benzoylecgonine in the acid extract (uncorrected for chemical conversion of cocaine to benzoylecgonine) ranged from 5.2 to 22.0 with a mean + SEM of 10.5 + 1.5, Ecgonine methyl ester ranged from 0.6 to 1.9 ng/mg. Norcocaine was assayed in 4 samples in amounts ranging from 0.5 to 0.7 ng/mg. Cocaethylene was measured in 6 samples in amounts ranging from 0.4 to 2.6 ng/mg. Traces of norcocaethylene were detected in those specimens containing cocaethylene. Ten drug-free control samples were processed in the same manner as the cocaine users' samples. The acid extracts from these samples were negative for cocaine and metabolites,
0 m
I
C~ III D) Sub. B / mW
Hair samples (ca. 100 mg) exposed to cocaine vapor were prewashed by two methods in an attempt to remove surface contamination. The amount of cocaine remaining on hair was measured by SIM at m/z 182; D3-cocaine was monitored at m/z 185 (Figure 3). Standard cocaine appeared at a retention time of 6.28 rain. (Figure 3A). As shown in Figure 3B, control hair exposed to drug-free room air was negative for cocaine. Samples exposed to cocaine vapor remained heavily contaminated with cocaine after washing with methanol (Figure 3C). The total amount of cocaine remaining on the hair samples ranged from 510-3,430 ng by the methanol procedure and 1,851-9,485 ng by the Baumgartner method (Table III). Washing with methanol ( 1 mL x 3) was effective in removing more than 90% of the cocaine from hair, whereas use of the alternate procedure reported by Baumgartner et al. (6) removed an average of 81%. Hair samples soaked in 1 mg/mL of aqueous cocaine hydrochloride remained grossly contaminated with cocaine after both wash procedures. No attempt was made to quantitate the amount of cocaine remaining in hair because all GC/MS responses were saturated (Figure 3D).
WASH
Discussion M/ZI~ --
It
Minutes
Figure 2. SIM recordings of standards, hair extracts, and methanolic wash for cocaethylene (CE), benzoylecgonine (BE), norcocaine (NC), and norcocaethylene (NCE). Panel A illustrates the response for a control hair extract containing 25 ng (total) of BE and 12.5 ng (total) of CE, NC, and NCE. Panels B and C illustrate the response for acid extracts of hair samples from Subjects B and A, respectively. Panel D represents the response for the methanolic wash prior to acid extraction of Subject B's hair sample.
Hair analysis for cocaine has been stated to be an efficient means of providing "... useful information on cocaine consumption by cocaine addicts ..." (2). Accordingly, Baumgartner reported finding a linear response between cocaine metabolite (presumably benzoylecgonine) in mouse hair and the administered dose (6). In the same article, he also reported finding a "reasonably good correlation" between cocaine metabolite in the hair of patients, probationers, and parolees and their self-reported amounts of cocaine use (6). In contrast, Kidwell (9) reported finding a large excess of intact cocaine over benzoylecgonine and ecgonine in the hair of users by tandem MS analysis under pyrolysis conditions. In the present study, we sought to clarify the nature of the chemical species that is excreted in hair
253
Journal of Analytical Toxicology, Vol. 15, September/October1991
same time or within a few hours (14). Hence, the detection of cocaethylene in 8 of 10 heavy cocaine users' hair is not surprising. The detection of traces of norcocaethylene, the Ndemethylated metabolite of cocaethylene, in the hair samples containing cocaethylene further supports the current finding of cocaethylene. The washing procedure was included in an attempt to eliminate potential surface contamination from environmental cocaine. Human head hair could be exposed to a variety of sources of cocaine free base and cocaine hydrochloride from the environment. Smoking "crack" involves the vaporization of base cocaine. Small amounts of "sidestream" crack vapor can be released into the atmosphere. This allows the possibility of adsorption of cocaine onto an individual's hair located nearby. Certainly, the smoker's hair would be contaminated in the same process. Alternately, cocaine hydrochloride dust particles in the air and on articles of clothing and money would present the opportunity for hair contamination from the powder and aqueous solutions. Although the environmentally contaminated controls utilized in this study simulated examples of extreme exposures from cocaine base and aqueous solutions of cocaine hydrochloride, other investigators have employed even more exaggerated conditions. Baumgartner and Hill (15) produced contaminated hair specimens for study by suspending locks of hair in a 1-L glass flask filled with vapor from 500 mg of cocaine freebase.
after cocaine use. This required the development of a sensitive and specific GC/MS assay for cocaine and metabolites for use with small amounts of hair samples. Controls were included to evaluate the potential role of environmental contamination of cocaine in hair analysis. Analysis of hair samples from the 10 documented cocaine users provided evidence for the presence of cocaine in each subject's hair, whereas negative findings were obtained for the 10 drug-free controls. The mean ratio of cocaine to benzoylecgonine was 10.5. Only traces of ecgonine methyl ester were detected. These findings agree with Kidwell's findings (9) that cocaine is the principal analyte in hair and occurs in much higher abundance than other cocaine metabolites. The detection of norcocaine and cocaethylene along with cocaine by GC/MS in the present study provided convincing proof that cocaine is excreted in hair and rules out the possibility of environmental contamination as an alternate explanation for cocaine's occurrence. Neither of these compounds occur in appreciable amounts in illicit cocaine and their presence can only be explained by human biotransformation of cocaine to these unique metabolites. Cocaine is oxidatively metabolized to norcocaine (12) and cocaethylene has been shown to arise as a result of the concurrent use of cocaine and alcohol (10). It is known that alcohol is often abused along with cocaine. A recent report indicates that approximately 85% of cocaine users polled in the 1985 National Survey on Drug Abuse used ethanol at the B) Control Vapor
A) Cocaine Std.
D) Aqueous Cocaine
C) Cocaine Vapor
;\
m/z 182
-
-
/
~-
m/z 185 6.'10
t 6.20
6.30
i 6.40
6.'10
6.20
6"30
m/z 182 m/z 185
6.40
61o
6.'~o 6.3o 6.40
,',0
6.'~0 6.30 ,.40
Minutes posed to room air and cocaine vapor (10 mg in a 2 x 3 x 3 ft box), respectively. Panel D illustrates the response for the acid extract (after methanol wash) for cocaine-free control hair soaked for 48 h in aqueous cocaine solution (1 mg/mL). Note the saturated response for cocaine in Panel D.
Figure 3. SIM recordings of a standard cocaine hair extract and environmentally contaminated hair samples. Panel A represents the response for standard cocaine (50 ng) at m/z182 (Rt = 6.29 rain) and deuterated cocaine (100 ng) at m/z185 (Rt = 6.28 min). Panels B and C illustrate the response for the acid extract (after methanol wash) for cocaine-free control hair ex-
Table III. Efficiency of Wash Procedures for Removal of Cocaine from Control Hair Specimens That Were Contaminated with Cocaine Vapor* Total CocaineRecovered(ng)
Percent cocaineremovedby washing
Wash fraction Wash Method Methanol
Ethanol-phosphate buffer (6)
Alcohol
Buffer
Acid extract fraction
Alcoholwash
Buffer wash
Total
11,052 30,954 118,320
-
510 1,019 3,430
95.6 96.8 97.2
-
95.6 96.8 97.2
9,238 31,160 t7,208
1,057 7,396 4,978
1,851 9,485 6,288
76.1 64.9 60.4
8.7 15.4 17.5
84.8 80.3 77.9
9 Triplicate drug-free hair specimens (Caucasoid, brown, ca. 100 mg each) were analyzed by two wash procedures as described in Methods section. The data indicate the total amount of cocaine recovered in ng by each procedure.
254
Journal of Analytical Toxicology, Vol. 15, September/October 1991
The methanol washing procedure used in the present study was more than 90% effective in removing vaporized cocaine, but sufficient cocaine residue remained to produce false positive results. After soaking in an aqueous solution of cocaine, the methanol wash procedure was totally inadequate in removing cocaine; hair samples remained extremely contaminated. The washing procedure reported by Baumgartner et al. (6) fared no better in removing cocaine contamination from these samples. Additional wash-out kinetic criteria have been proposed by Baumgartner and Hill (15) as a means of distinguishing external cocaine contamination from active drug use; however, these criteria remain untested by other investigators. Other studies have included a variety of conditions in attempts to remove drug contamination. Marigo et al. (16) reported that serial washings with ethyl ether and dilute HC1 successfully removed "loosely bound morphine." In contrast, Kidwell (9) reported the failure of a variety of wash solutions in removing phencyclidine from contaminated hair. Analysis o f the methanolic wash for cocaine metabolites produced evidence that the wash procedure removed more than just external cocaine contamination. The presence of norcocaine and cocaethylene along with cocaine and benzoylecgonine in amounts generally exceeding those found in the extract indicated that interior drug components were removed in the wash. Clearly in the present procedure, the "wash" procedure could more aptly be called an extraction. The depth of penetration by the "wash" solution is likely to be dependent upon the condition of the hair undergoing testing. Highly damaged hair is likely to be more porous and yield greater amount of interior drug during washing. If this is true, the amount of drug remaining in the extract is likely to be distorted and bear little relationship to dosing patterns. Based on the present data, it seems unlikely that there is a clear demarcation in hair between environmental drug exposure and drug residues resulting from active use. The elimination of surface contamination from hair prior to its analysis for drugs of abuse is important, given the possibilities of environmental contamination. Unfortunately, in the present effort, the techniques employed were not effective. It may be necessary to adopt an alternate approach in the development of hair analysis technology by requiring the identification of a unique metabolite whose presence can only be explained by internal drug exposure with subsequent biotransformation. The presence of cocaine, benzoylecgonine, and ecgonine methyl ester could potentially be explained by environmental contamination with subsequent hydrolysis. In sharp contrast, the presence of either norcocaine or cocaethylene in an individual's hair would appear to constitute sufficient evidence of internal cocaine exposure.
References 1. M. Saitoh, M. Uzuka, and M. Sakamoto. Rate of hair growth. In Advances in Biology of Skin. Vo/. IX Hair Growth. W. Montagna and R.L Dobson, Eds., Pergamon, Oxford, pp. 183-201 (1969). 2. D. Valente, M. Cassini, M. Pigliapochi, and G. Vansetti. Hair as the sample in assessing morphine and cocaine addiction. Olin. Chem. 27:1952-53 (1981). 3. W.A. Baumgartner, C.T. Black, P.F. Jones, and W.H. Blahd. Radioimmunoassay of cocaine in hair: Concise communication. J. NucL Med. 23." 790-92 (1982). 4. F.P. Smith and R.H. Liu. Detection of cocaine metabolite in perspiration stain, menstrual bloodstain, and hair. J. Forens. Sci. 31." 1269-73 (1986). 5. S.A. Reuschel and F.P. Smith. Benzoylecgonine (cocaine metabolite) detection in hair samples of jail detainees using RIA and GC/MS. J. Forens. Sci., in press, 1991. 6. W.A. Baumgartner, V.A. Hill, and W.D. Blahd. Hair analysis for drugs of abuse. J. Forens. Sci. 34:1433-53 (1989). 7. S. Balabanova and J. Homoki. Determination of cocaine in human hair by gas chromatography/mass spectrometry. Z. Rechtsmed. 98:235-40 (1987). 8. RM. Martz. The identification of cocaine in hair by GC/MS and MS/MS. Crime Lab. Digest 15:67-73 (1988). 9. D.A. Kidwell. Analysis of drugs of abuse in hair by tandem mass spectrometry. In Proceedings, 36th American Society of Mass Spectrometry Conference on Mass Spectrometry and Allied Topics, San Francisco, CA, 1989, pp. 1364-65. 10. W.L Hearn, D.D. Flynn, G.W. Hime, S. Rose, J.C. Cofino, E. Mantero-Atienza, C.V. Wetli, and D.C. Mash. Cocaethylene: A unique metabolite displays high affinity for the dopamine transporter. J. Neurochem. 56:698-701 (1991). 11. Personal communication, W. Lee Hearn. 12. T. Inaba, D.J. Stewart, and W. Kalow. Metabolism of cocaine in man. Clin. Pharmacol. Ther. 23:547-52 (1978). 13. C.E. Cook, A.R. Jeffcoat, and M. Perez-Reyes. Pharmacokinetics studies of cocaine and phencyclidine in man. In Pharmacokinetics and Pharmacodynamics of Psychoactive Drugs, G. Barnett and C.N. Chiang, Eds., Biomedical Publications, Foster City, CA, 1985, pp. 48-74. 14. B.F. Grant and T.C. Harford. Concurrent and simultaneous use of alcohol with cocaine: Results of national survey. Drug Alcohol Dep. 25:97-104 (1990). 15. W.A. Baumgartner and V.A. Hill. Hair analysis for drugs of abuse: Decontamination issues. In Proceedings of the 2rid International Congress of Therapeutic Drug Monitoring and Toxicology, Oct. 9-12, 1990, Barcelona, Spain, in press. 16. M Marigo, F. Tagliaro, C. Poiesi, S. Lafisca, and C. Neri. Determination of morphine in the hair of heroin addicts by high performance liquid chromatography with fluorimetric detection. J. Anal. ToxicoL 10:158-61 (1986).
255
