Highly Sensitive Micro-Plate Enzyme Immunoassay Screening and ...

Journal of Analytical Toxicology,Vol. 23, October 1999

Highly SensitiveMicro-Plate Enzyme Immunoassay Screeningand NCI-GC-MS Confirmation of Flunitrazepam and Its Major Metabolite 7-Aminoflunitrazepam in Hair'* Adam Negrusz 1,~,Christine Moore 2, Dawn Deitermann 2, Douglas Lewis2, Kristen Kaleciak 1, Robert Kronstrand 3, Brian Feeley4, and R. Sam Niedbala 4 I Department of Pharmaceutics and Pharmacodynamics,College of Pharmacy, University of Illinois at Chicago, 833 South Wood Street, Chicago, Illinois 60612; 2United States Drug TestingLaboratories, Inc., 1700 South Mount Prospect Road, Des Plaines, Illinois 60018; 3National Board of Forensic Medicine, Link6ping, Sweden; and 4STCTechnologies Inc., 1745 EatonAvenue, Bethlehem, Pennsylvania 18018

Abstract Flunitrazepam (Rohypnol) is a benzodiazepine used in the treatment of insomnia as a sedative hypnotic and as preanesthetic medication in European countries and Mexico. Although it has no medicinal purpose in the United States, the occurrence of its abuse is increasing. Sexual abuse of both men and women while under the influence of so-called "date-rape" drugs has been the focus of many investigations. Reported date-rape drugs include flunitrazepam (FN), clonazepam, diazepam, oxazepam, gammahydroxybutyrate, and many others. FN has been banned in the United States because of its alleged use in such situations. Unfortunately, the detection of FN or its metabolites 7aminoflunitrazepam (7-AFN) and desmethylflunitrazepam in a single specimen such as urine or blood is difficult in criminal situations because of the likelihood of single-dose ingestion and the length of time since the alleged incident. Hair provides a solution to the second of these problems in that drugs tend to incorporate into hair and remain there for longer periods of time than either urine or blood. There are various techniques for the detection of FN in plasma, blood, and urine, but little work has been done with hair. Hair collection is a virtually noninvasive procedure that can supply information on drug use for several months preceding collection. The objective of this paper was to determine if a commercially available micro-plate enzyme immunoassay system was sufficiently sensitive for the routine screening of 7-AFN in hair by the development of extraction procedures and optimization of the immunoassay kit. Further, this study used the same solid-phase extraction to isolate FN and its major metabolite, 7-AFN, and gas chromatography-mass spectrometry with negative ion chemical ionization for confirmation. Two seven-point standard curves were established ranging from 0.5 pg/mg to 100 pg/mg for 7-AFN and 2.5 pg/mg to * This paperwas presentedat the joint 1998 meeting of the International Association of Forensic Toxicologists (TIAFT)and Society of ForensicToxicologists in Albuquerque, New Mexico. f Author to whom correspondenceshould be addressed.

200 pg/mg for FN with respective deuterated internal standards. A replicate analysis of controls was performed to establish interand intraday variabilities. Two suicide cases along with one alleged date-rape case and one case of an emergency room patient whose blood screened positive for benzodiazepines were analyzed. All the hair specimens screened positive for benzodiazepines using micro-plate enzyme immunoassay. Two cases, including the daterape case, were negative for FN and 7-AFN, and two postmortem hair samples were confirmed positive for FN and its metabolite.

Introduction Flunitrazepam (Rohypnol or "Roofies") is a benzodiazepine [5-(2-fluorophenyl)-l,3 dihydro-l-methyl-7-nitro-l,4-benzodiazepine-2-one] with a very strong hypnotic effect in addition to the sedative, anxiolytic, muscle relaxant, and anticonvulsant effects characteristic of other benzodiazepines (1). It is approximately 10 times more potent than diazepam (Valium| and is used outside of the U.S. for treatment of insomnia and as a premedication in anesthesiology. Therapeutic doses range from 1 to 2 mg. Flunitrazepam is used in Europe and Mexico for the treatment of sleep disorders. However,sexual abuse of both men and women while under the influence of this so-called "date-rape" drug has been the focus of many investigations. In response to this, flunitrazepam was recently banned as a prescription drug in the U.S. Following administration of a single oral dose of 2 or 4 mg of Rohypnol, flunitrazepam and its metabolites 7-aminoflunitrazepam and desmethylflunitrazepam have been detected in plasma and urine up to 60 h postdose at measurable concentrations (2) and up to 72 h by analysis of the related benzophenones (3). Administered doses of 1 mg did not result in

Reproduction(photocopyin8)of editorialcontentof thisjournalis prohibitedwithoutpublisher'spermission.

429

Journal of Analytical Toxicology, Vol. 23, October 1999

positive urine or plasma. Rape victims do not alwaysreport the alleged incident in a timely manner because of fear, shame, disbelief, and other psychological reasons. It is not unusual for a urine or blood analysis to provide a negative result because of the likelihood of a single, low-dose ingestion, urine dilution, and the length of time since the alleged incident. Hair provides a solution to this historical aspect of testing in that basic drugs such as flunitrazeparn and 7-aminoflunitrazepam tend to incorporate into hair and remain there for longer periods of time than either urine or blood. The detection of benzodiazepines in hair has not been widely reported. Only radioimmunoassay techniques have addressed the screening of hair samples for benzodiazepines (4), and other papers have identified lorazepam and desmethyldiazepam using negative ion gas chromatography-mass spectrometry (NCI-GC-MS)(5). In another study, Cirimele et al. (6) screened hair samples for several benzodiazepines,including flunitrazepam, norcliazepam, oxazepam, bromazepam, diazepam, lorazepam, alprazolam, and triazolam, using NCI-GC-MS. Out of 115 hair specimens collected, flunitrazepam was present in 31, and the concentration range was 19-148 pg/mg. In a more recent study, Yegleset al. (7) reported concentrations of several benzodiazepines, including flunitrazepam, in hair samples obtained postmortem from 21 individualswho died from overdoses of legal and illicit drugs. Maximal flunitrazepam concentrations were as high as 1.8 to 9.5 ng/mg. It has been reported that following ingestion of flunitrazepam, it is the rnetabolite 7-aminoflunitrazepam that appears in greater concentrations in hair than the parent drug itself, and in 12 out of 40 cases, 7-aminoflunitrazepamwas the only compound found (8). Of 40 hair samples, only 14 were positive for both flunitrazepam and 7-aminoflunitrazepam, and concentrations ranged from 31 to 129 pg/mg for flunitrazepam (14 cases) and from 3 pg/mg to 161 pg/mg for 7-aminoflunitrazepam (26 cases). The same study reported detection limits of 15 and 3 pg/mg for flunitrazepam and 7-aminoflunitrazepam, respectively,using NCI-GC-MS methodology.An earlier publication (9) identified parent flunitrazepam and 7-aminoflunitrazepam in the hair of a Rohypnol abuser at concentrations of 89.5 and 24 ng/mg, respectively.However,to date, there are no published screening procedures for the determination of 7aminoflunitrazepam in hair. The first objective of this paper was to determine if a commercially available micro-plate enzyme immunoassay system was sufficiently sensitive for the routine screening of 7aminoflunitrazepam in hair. The second goal was to develop and validate a highly sensitive confirmatory NCI-GC-MS method to followsolid-phase extraction (SPE) of flunitrazepam and 7-aminoflunitrazepam from the hair matrix. We also intended to apply both screening and confirmation to cases involving flunitrazepam use.

Experimental Micro-plate enzyme immunoassay Materials. Mixed-mode SPE columns (200 rag, 10 mL, Isolute| HCX)were obtained from Jones Chromatography (Lakewood, CO). The 7-aminoflunitrazepam standard was obtained 430

from Radian Corp. (Austin, TX).STC Technologies,Inc. (Bethlehem, PA) generously loaned the microplate washer and reader and supplied Benzodiazepine Urine Micro-Plate kits and Stabilzyme buffer. Glacial acetic acid, methanol, hydrochloric acid, methylene chloride, isopropanol, and concentrated ammonium hydroxide were purchased from Fisher Scientific (Itasca, IL),and they were of high-performance liquid chromatography (HPLC) grade or better. Sample preparation. Drug-free hair was powdered and weighed out into 50-rng aliquots. The 7-aminoflunitrazeparn standard was added to the hair at concentrations of 0.05, 0.1, 0.2, 0.5, and 1.0 ng/mg. Methanol (3 mL) was added, and the specimens were sonicated for 5 h. The methanol was decanted, 0.1N hydrochloric acid (3 mL) was added to the hair, and the specimens were incubated overnight at 55~ The following day, the methanol and acid were combined, and 1.93M glacial acetic acid (1 mL) and deionized water (9 mL) were added. Solid-phaseextraction. Mixed-modeSPE columns (200 mg, 10 mL, Isolute HCX)were conditioned with methanol (3 mL), deionized water (3 mL), and 1.93M acetic acid (1 mL). The sample was added to the column through a filter and drawn through the sorbent bed slowly. The bed was washed with deionized water (3 mL), 0.1N hydrochloric acid (1 mL), and methanol (3 mL). Collection tubes were placed into the rack, and the basic drugs were eluted in methylene chloride/isopropanol/ammonium hydroxide (78:20:2, v/v, 3 mL). The eluent was evaporated to dryness, and the residues were reconstituted in 50% Stabilzyme buffer (100 pL). Method. The specimens were screened using a benzodiazepine urine micro-plate enzyme immunoassay kit. In order to optimize kit performance, the reconstituted sample, calibrator, or control (50, 25, or 10 pL) was added to each well of the microplate. All specimens were tested in duplicate. The enzyme conjugate (100 pL) was added, and the plate was incubated at room temperature in the dark for 1, 1.5, 2, or 3 h to further optimize the system. The plate was emptied and washed with deionized water (6 x 300 pL), and substrate reagent was added (100 pL). The plate was again incubated in the dark, at room temperature, for 30 min as recommended by the manufacturer. After incubation, stopping reagent was added (100 pL) and the absorbance was read using a microplate reader. Calibrators supplied with the Benzodiazepine Urine Assay were used with each plate serving as a means of quality control of the kit. Extracted hair spiked at concentrations of 0.05, 0.1, 0.2, 0.5, and 1 ng/mg, as well as spiked-positive hairs (0.5 ng/mg) and negative hair, were run with each plate to allow for matrix effects.

Chemical ionization GC-MS Instrumentation. The GC-MS system consisted of a Hewlett Packard 6890 series injector, an HP 6890 series GC system, and an HP 5973 mass selective detector with positive and negative ion chemical ionization capabilities (Hewlett Packard, Palo Alto, CA). An HP-5MS capillary column (30 m x 250 pm x 0.25 pro) was used for separation (Hewlett Packard, Palo Alto, CA).The hair pulverizer was from Crescent (Lyons, IL), and the heating block was from Fisher Scientific.The Vac-Elut extraction manifoldwas from AnalyticalInternational (Varian,Harbor City, CA),the centrifuge from Damon/IECDivision (Needham


Heights, MA), and the Meyer N-EVAP| analytical evaporator from Organomation Assoc., Inc. (Northborough, MA). Chemicals, reagents, and specimens. Flunitrazepam (1 mg/mL in methanol), 7-aminoflunitrazepam (100 pg/mL in acetonitrile), and the deuterated internal standards flunitrazepam-d7 (100 pg/mL in methanol) and 7-aminoflunitrazepam-d7 (100 pg/mL in acetonitrile) were all purchased from Radian International (Austin, TX). Methanol (HPLC grade), hydrochloric acid (certified ACS Plus), glacial acetic acid (HPLCgrade), methylene chloride (HPLC/GC-MSgrade), isopropanol (HPLC grade), ethyl acetate (HPLC grade), and concentrated ammonium hydroxide (certified ACS Plus) were bought from Fisher Scientific. Heptafluorobutyric anhydride (HFBA)was purchased from CampbellSupply Co. (Rockton,IL). The HCX Isolute columns (International Sorbent Technologies) were purchased from Jones Chromatography (Lakewood, CO). The four collected hair samples screened positive for benzodiazepines. Specimen A was obtained from an alleged rape victim from the Chicagoarea; specimen B was obtained from an emergency room patient whose blood screened positive for benzodiazepines; and specimens C and D were collected from corpses and sent for analysis by the National Board of Forensic Medicine (Link~ping,Sweden).The concentrations of all other drugs in femoral blood were determined using liquid-liquid extraction and Hewlett Packard 5880A GC equipped with a nitrogen-phosphorus (NPD)detector. Further information about the postmortem cases was as follows.Specimen C was obtained from a 70-year-old,grey-haired male who had committed suicide with prescription drugs. His stomach contents contained tablet residues, and his femoral blood contained the following substances: dextropropoxyphene(1.2 pg/g), flunitrazepam (0.02 pg/g), 7-aminoflunitrazepam (0.3 lJg/g), citalopram (0.3 pg/g), desmethylcitalopram(0.1 IJg/g),and acetaminophen (140 pg/g). There was only 9 mg of grey hair available for NCI-GC-MS confirmation. Specimen D was obtained from a 33-year-old female who also committed suicide with prescription drugs. The hair was apparently bleached a few months before death. Her femoral blood contained the followingsubstances: carisoprodol (0.4 ~g/g), dextropropoxyphene(0.4 pg/g), trimipramine (0.2 pg/g), chlormezanone (23 IJg/g), diazepam (0.6 1Jg/g), nordiazepam (0.2 ~g/g), flunitrazepam (0.02 pg/g), 7-aminoflunitrazepam (0.2 ~g/g), 7-aminonitrazepam (0.4 pg/g), and citalopram (1.2 pg/g). The analysis of human specimens including hair was approved by the Institutional ReviewBoard (# 97-996) of the Universityof Illinois at Chicago. All samples, including negative hair, were washed in deionized water, dried, pulverized, and stored at room temperature before analysis. Standards and controls. The flunitrazepam (1 mg/mL in methanol) standard stock solution was diluted to 100 pg/mL, 10 pg/mL, 1 pg/mL, and 0.2 IJg/mL.The 7-aminoflunitrazepam (100 pg/mL in acetonitrile) standard stock solution was diluted to 10 pg/mL, 0.1 pg/mL, and 10 ng/mL. The flunitrazepam-d7 (100 ~g/mL in methanol) and 7-aminoflunitrazepam-d7 (100 pg/mL in acetonitrile) deuterated internal standards were diluted to 10 ~g/mL and 100 ng/mL. All standards were diluted in their respective solvents. Seven-point standard curves were made for both fluni-

trazepam and 7-aminoflunitrazepam by using a 50-rag aliquots of the negative pulverized hair spiked with solutions of both drugs. The concentrations of the flunitrazepam in standard hair preparations were as follows:2.5, 5.0, 10.0, 20.0, 50.0, 100.0, and 200.0 pg/mg of hair. The concentrations of 7-aminoflunitrazepam were 0.50, 1.0, 5.0, 10.0, 20.0, 50.0, and 100.0 pg/mg of hair. In addition, two levelsof controls were prepared. The low controls for flunitrazepam (15 pg/mg) and 7-aminoflunitrazepam (3 pg/mg) were prepared by adding 75 pL of 10 pg/pL flunitrazepam and 15 pL of 10 pg/IJL 7-aminoflunitrazepam, respectively, to 50-rag aliquots of negative hair. The high controis for flunitrazepam (80 pg/mg) and 7-aminoflunitrazepam (40 pg/mg) were prepared by adding 20 pL of 200 pg/pL flunitrazepam solution and 20 IJL of 100 pg/pL 7-aminoflunitrazepam solution, respectively. Analyticalprocedure.The hair samples were pulverized and 50-rag aliquots were analyzed.In one case, only 9 mg of hair was available for analysis. To the hair samples, standard and control preparations, 50 pL of 100 pg/pL of flunitrazepam-d7and 10 mL of 100 pg/1JLof 7-aminoflunitrazepam-d7were added. Methanol (3 mL) was added to the hair, and it was sonicated for 1 h. The methanol layer was decanted and stored in the refrigerator. Hydrochloric acid (0.1N, 3 mL) was added to the hair, and the specimens were incubated overnight at 55~ after being sealed and vortex mixed. The test tubes were removed from the heating block and centrifugedfor 5 min at 400 • The methanol and acid were combined, and 1.93M glacial acetic acid (1 mL) and deionized water (9 mL) were added. Mixedmode Isolute HCXSPE columns were conditioned with the following and not allowed to dry: methanol (3 mL), deionized water (3 mL), and 1.93M glacial acetic acid (1 mL). The sample was then added to the column and drawn through slowly.The column was allowedto dry for approximately2 rain. The bed of the column was washed with 3 mL of deionizedwater (dried for 1-2 rain), 1 mL of 0.1N hydrochloric acid (dried for 1-2 rain), and 3 mL of methanol (dried for 5 rain). The collection tubes were placed in the rack, and the drugs were eluted using mixture of methylene chloride/isopropanol/ammonium hydroxide (78:20:2, v/v/v).The eluent was then evaporatedto dryness using a stream of nitrogen. The dry residue was reconstituted in 50 pL of ethyl acetate and transferred to autosampler vials. The extract was evaporated to dryness in the vacuum oven at 60~ The samples were derivatized using 50 pL of HFBA at 60~ for 30 rain in the sealed vials. After incubation, the derivatizing agent was evaporated in the vacuum oven (60~ and the dry residue reconstituted in 25 ~L of ethyl acetate. Chromatographicmethod. The injector was operated in the splitless mode, and the injection volume was 1 ~L. The injector temperature was 240~ Ultra-high-purity helium (99.999%) was used as the carrier gas at a flow rate of 1.2 mL/min. The initial GC oven temperature of 60~ was held for I min, and then increased at a rate of 30~ until the final temperature of 310~ was attained. The final temperature was held for 3 min. The total run time for one injection was 12.33 min. Methane (ultra-high purity, 99.99%) was used as reagent gas at an apparent pressure of 3.7 • 10-4 Torr in the ion source (methane flow 3.25 mL/min.) The MS ion source temperature was 250~ and the quadrupole temperature was 106~ The electron mul431


tiplier voltage was set at +400V above the NCI-tune voltage. The mass selective chemical ionization detector monitored negative ions (NCI)and was operated in the selected ion monitoring (SIM) mode. The solvent delay was 9 rain. The following ions were monitored and used for quantitation: m/z 313 for flunitrazepam, m/z 320 for flunitrazepam-dT,m/z 459 for 7-arninoflunitrazepam, and m/z 466 for 7-aminoflunitrazepamdT.The dwell times were 20 ms for m/z 313 and 459 and 50 ms for m/z 320 and 466. In negative chemical ionization mode, a single specific ion is indicative of the presence of a molecule, particularly when coupled with other supporting information such as retention time and screening data. It was necessary to change the liner frequently because flunitrazepam is particularly sensitive to active sites. Precision and accuracy. Quantitation of flunitrazepam and 7-aminoflunitrazepam was performed by the internal standard method. A seven-point standard curve was prepared by linear least-square regression analysis of the ratio of the peak area of flunitrazepam to the peak area of the internal standard, flunitrazepam-dT. A separate seven-point standard curve was also prepared for 7-aminoflunitrazepam with 7-aminoflunitrazepam-d7 as the internal standard. Peak-area ratios were determined for the controls. Control concentrations were calculated from the standard curve values. Intraday variability was ascertained by analyzing four replicates of low controls (15 pg/mg and 3 pg/mg) and high controls (80 pg/mg and 40 pg/mg) for flunitrazepam and 7-aminoflunitrazepam, respectively. Interday variability was ascertained over a period of four weeks. The mean measured concentrations and standard deviations were calculated based on the inter- and intraday variability populations. The percent relative accuracy was calculated by the following equation: [(Mean Measured Concentration -Theoretical Concentration) / Theoretical Concentration] x 100%. All data were acquired and analyzed by HP software, Enhanced G1701BA ChemStation version B.00.00 for Windows NT version 4.0.

munoassay kit itself is targeted towards oxazepam. The crossreactivity of structurally related compounds, including 7aminoflunitrazepam, is availablein the package insert provided by the manufacturer, and the cross-reactivity towards 7-aminoflunitrazepam is 156% oxazepam equivalents. The optimum incubation time for plates was determined to be 1 h and the optimum sample size 50 IJL because the greatest difference in absorbance values between drug-free hair and spiked hair was observed using those parameters. The final assay allowed a sufficient difference between absorbance produced by the negative hair and that produced by hair containing "I|

A

t

I q

B

e~

the

e~

aline ~

e.h)

' .

,

9

Results Micro-plate enzyme immunoassay The suggested application of this assay is for use in sexual assault situations where the use of flunitrazepam is suspected, even though the benzodiazepine micro-plate enzyme imTable I. Micro-Plate Enzyme Absorbance Values for Hair Specimens

Concentrationequivalents 7-aminoflunitrazepam Specimen type "Positive" hair Negative hair A (rape victim) B (emergency room) C (postmortem) D (postmortem)

432

Mean absorbance 0.598 (n = 20) 2.353 (n = 20) 1.147 0.687 1.504 0.093

(n~/m~) 0.5 0 0.2 0.5 0.1 >1

Figure 1. Selected ion chromatograms (A, m/z 466; B, m/z 459; C, m/z 320; and D, m/z 313) of the control hair preparation (7-aminoflunitrazepam concentration 3 pg/mg, flunitrazepam concentration 15 pg/mg).


0.05 ng/mg to determine positivity; however, a cutoff concentration of 0.1 ng/mg was selected for use because hair specimens may contain flunitrazepam as well as 7-aminoflunitrazepam or other related metabolites that may cause cross-reactivity. The procedure was validated using drug-free negative hairs (n = 20) and "positive" hairs (n = 20) spiked at 0.5 ng/mg 7-aminoflunitrazepam. The negative hairs averaged an absorbance value of 2.353 ; the "positive" hairs averaged an absorbance value of 0.598. The difference in absorbance values was sufficient to allow identification of positive hairs using a cutoff level of 0.1 ng/mg. All results are presented in Table I. U

A

lm

,lleeoo

Z 1~ooo

sl.a~

Chemical ionization GC-MS

Figure I presents the selected ion chromatograms (m/z 466, m/z 459, rn/z 320, m/z 313) of the extract from blank hair spiked with 7-aminoflunitrazepam, flunitrazepam (3 pg/mg and 15 pg/mg, respectively), and corresponding deuterated internal standards. Figure 2 shows the selected ion chromatograms (m/z 466, m/z 459, m/z 320, m/z 313) for the postmortem sample (sample C). In this case, only 9 mg of the grey hair was available for NCI-GC-MS confirmation. Selected ion chromatograms (m/z 320 and rn/z 313) for the second postmortem sample analyzed (sample D) are presented in Figure 3. Flunitrazepam concentration was 23 pg/mg. The retention time of 7-aminoflunitrazepam and 7-aminoflunitrazepam-d? was 9.56 min, and the retention time of flunitrazepam and flunitrazepam-d7 was approximately 9.99 rain. All chromatograms were recorded over the time range of 9.00 to 11.00 min. Standard curves for flunitrazepam and 7aminoflunitrazepam were linear over the range of drugs assayed (0.5 pg/mg to 100 pg/mg for 7-aminoflunitrazepam and 2.5 pg/mg to 200 pg/mg for flunitrazepam) and had correlation coefficients 0.998 and 0.997, respectively. The limit of quantitation for 7-aminoflunitrazepam was 0.5 pg/mg and for flunitrazepam 2.5 pg/mg for 50 mg of sample. Both limits of quantitation were arbitrarily established to be the lowest concentrations on the corresponding standard curves. The limits of detection were 1.5 pg/mg for flunitrazepam and 0.2 pg/mg for 7-aminoflunitrazepam and were the lowest concentrations of drugs at which the signal-to-noise ratio was 3:1. Tables II and III present the accuracy and precision of the flunitrazepam and 7-aminoflunitrazepam control hair preparations, respectively. The intraday variability was determined by

Jzvn*lm

m

:=1

T

,.

o

14om

"t|

B

lzoee i1 iooeo

,il

Figure 2. Selected ion chromatograms (A, m/z466; B, m/z 459; C, m/z 320; and D, m/z 313) of the postmortem sample C (9 mg available for confirmation).

Figure 3. Selected ion chromatograms (A, m/z 320 and B, m/z 313) of the postmortem sample D (flunitrazepam concentration 23 pg/mg).

433


analyzing four replicates of controls (15 and 80 pg/mg of hair for flunitrazepam and 3 and 40 pg/mg of hair for 7-aminoflu~ nitrazepam) prepared in hair on a single day. The interday variability was determined over a period of four weeks on eight separate days. Table IV presents the results of quantitation of flunitrazepam and 7-aminoflunitrazepam in four hair samples (one collected in the emergency room, one from the alleged rape victim from Chicago area, and two postmortem cases). Only hair specimens collected from corpses (suicides using prescription drugs) were

Table II. Accuracy and Precision of Flunitrazepam Hair Preparations (pg/mg) Parameter

low control

High control

Theoretical concentration 15 pg/mg 80 pg/mg Intraday variability N=4 N=4 Mean measured 15.86 (• 0.75) pg/mg 71.12 (• 5.82) pg/mg concentration (• S.D.) % Coefficient of variation 4.73 8.18 % Relative accuracy 5.73 -11.10 Interday variability N = 13 N = 13 Mean measured 14.58 (• 1.31) pg/mg 70.40 (• 4.82) pg/mg concentration (r S.D.) % Coefficient of variation 8.98 6.85 % Relative accuracy -2.80 -12.00

Table III. Accuracy and Precision of 7-Aminoflunitrazepam Hair Preparations (pg/mg) Parameter

tow control

Highcontrol

Theoretical concentration

3 pg/mg N=4 2.81 (• 0.26) pg/mg

40 pg/mg N=4 36.31 (• 3.72) pg/mg

9.25 -6.33 N = 14 2.93 (+ 0.28) pg/mg

10.25 -9.23 N = 14 38.21 (+ 3.39) pg/mg

9.56 -2.33

8.87 -4.48

Intraday variability Mean measured concentration (• S.D.) % Coefficient of variation % Relative accuracy

Interday variability Mean measured concentration (• S.D.) % Coefficient of variation % Relative accuracy

Table IV. Flunitrazepam and 7-Aminoflunitrazepam Concentrations in Hair Samples Analyzed Sample A (rape victim) B (emergency room) C (postmortem)~ D (postmortem)

Flunitrazepam (pg/mg) N.D.* N.D. 1.76* 23.00

* not detected. greyhair,9 m8 availablefor analysis. * belowquantitationlimit.

434

7-Aminoflunitrazepam (pg/mg) N.D. N.D. 48.60 26.10

positive for both flunitrazepam and its major metabolite 7aminoflunitrazepam. In one case (9-rag sample), the flunitrazepam concentration was below the quantitation limit.

Discussion During the last several years, a notable increase of the drugfacilitated sexual assaults has been observed (10); however, to date, no complete and current statistical data are available. In the early 1990s, flunitrazepam was frequently used as a daterape drug and for that reason has been banned in the United States. In a recent study, Rickert et al. (11) presented results of the cross-sectional survey on 904 sexually active women 14 to 26 years of age. Of the 904 women, 5.9% admitted using flunitrazepam voluntarily one or more times; 7.5% of those who used flunitrazepam reported using it 10 to 19 times, and 11% reported using it 20 or more times. In addition, 10% of voluntary users reported subsequent physical or sexual victimization. Drug-facilitatedsexual assault victims do not always report the alleged incident in a timely manner because of fear, shame, disbelief, and other psychological reasons. Blood and urine analysis in such cases usually provide a negative result because of the likelihood of a single low-dose ingestion and the length of time since the alleged incident. A few years ago, hair became a possible matrix to evaluate exposure to drugs because it provides a solution to this historical aspect of testing, in that basic drugs (12) such as flunitrazepam and 7-aminoflunitrazepam tend to incorporate into hair and remain there for a long period of time. We suggest that if sufficient time is allowed to elapse between the alleged assault and collection of the sample (approximatelyseven days),hair can be used for analysis. Radioimmunoassay techniques have addressed the screening of hair samples for benzodiazepines (4). However,to date, there are no published screening procedures for the determination of 7-aminoflunitrazepam in hair. In our study, the sensitivity of the screening assay was not as good as the NCI-GC-MS limits reported (6--8), but all of the specimens, including the grey and bleached hairs from the postmortem victims, screened positive for benzodiazepines using the described methodology (Table I). Presumably, the subjects from whom specimens C and D were taken were regular users of benzodiazepines. In fact, hair specimen D was taken from the head of a person whose femoral blood had already tested positive for benzodiazepines other than flunitrazepam and 7-aminoflunitrazepam. The presence of flunitrazepam and 7-aminoflunitrazepam was confirmed using NCI-GC-MS in two out of the four samples (C and D). Specimens A and B did not confirm for flunitrazepam or its metabolite, but were positive using the screening assay. The presence of other benzodiazepines (e.g., oxazepam, diazepam, clonazepam) is a possible explanation of these results. Samples C and D screened positively for benzodiazepines at concentrations of 1.76 and 48.60 pg/mg (sample C) and 23.00 and 26.10 pg/mg (sample D) of flunitrazepam and 7-aminoflunitrazepam, respectively. Similar NCI-GC-MS analytical methodology for flunitrazepam and 7-aminoflunitrazepam was previously reported and applied to hair analysis


by Cirimele and co-workers (8,9) with solvent extraction and diazepam-d5as an internal standard. In these studies, limits of detection for flunitrazepam and 7-aminoflunitrazepamwere 15 and 3 pg/mg, respectively,and the limit of quantitation was 10 pg/mg for both compounds. Our study significantly improves the quantitation limits for flunitrazepam (2.5 pg/mg) and for 7-aminoflunitrazepam (0.5 pg/mg). All precision and accuracy data were within acceptable limits (Tables II and III). The vast majority of analytical methods require 50 mg of hair for analysis. In addition, concentrations of many drugs in grey hair are significantly lower than in pigmented hair (13). In this study we were able to succesfully apply SPE and NCI-GC-MS to detect and quantitate flunitrazepam and its major metabolite 7aminoflunitrazepam in 9 mg of grey hair. Even though the confirmed-positive hair specimens were taken from benzodiazepine users, we hypothesize that the sensitivity of the described methods is sufficient for the determination of much lower doses of flunitrazepam use, possibly following single administrations. Further research is ongoing at this time.

Conclusions The STC Micro-Plate EIA benzodiazepine system described was sufficiently sensitive for the screening of 7-aminoflunitrazepam in hair at a cutoff concentration of 0.1 ng/mg. This concentration appeared to produce positive screening results from four hair specimens suspected to contain benzodiazepines. The analytical method presented in this paper for the determination of flunitrazepam and its major metabolite 7-aminoflunitrazepam in hair employing SPE and NCI-GC-MS is extremelysensitiveand allowsaccurate and precise quantitation of both drugs in hair samples collected in various forensic cases.

Acknowledgments This project was partially supported under Awardnumber 98LB-VX-K020from the Office of Justice Programs, National Institute of Justice, Department of Justice. Points of view in this paper are those of the authors and do not necessarily represent the officialposition of the U.S. Department of Justice.

References 1. M.A.K. Mattila and H.M. Lami. Flunitrazepam: a review of its pharmacological properties and therapeutic use. Drugs 20: 353-374 (1980). 2. S.J. Salamone, S. Honasoge, C. Breener, A.J. McNally, J. Passarelli, K. Goc-Szkutnicka, R. Brenneisen, M.A. EISohly, and S. Feng. Flunitrazepam excretion patterns using the Abuscreen OnTrak and OnLine immunoassays: comparison with GC-MS. J. Anal. Toxicol. 21:341-345 (1997). 3. M.A. EISohly, S. Feng, S.J. Salamone, and R. Wu. A sensitive GC-MS procedure for the analysis of flunitrazepam and its metabolites in urine. J. Anal. Toxicol. 21" 335-340 (1997). 4. J.J. Sramek, W.A. Baumgartner, T.N. Aherns, V.A. Hill, and N.R. Cutler. Detection of benzodiazepines in human hair by radioimmunoassay. Ann. Pharrnacother. 26:469-471 (1992). 5. P. Kintz, V. Cirimele, F. Vaysette, and P. Mangin. Hair analysis for nordiazepam and oxazepam by gas chromatography-negative-ion chemical ionization mass spectrometry. ]. Chromatogr. 677" 241-244 (1996). 6. V. Cirimele, P. Kintz, and B. Ludes. Screening for forensically relevant benzodiazepines in human hair by gas chromatography-negative ion chemical ionization-mass spectrometry. J. Chrornatogr. BT00:119-129 (1997). 7. M. Yegles, F. Mersch, and R. Wennig. Detection of benzodiazepines and other psychotropic drugs in human hair by GC/MS. Forensic Sci. Int. 84" 211-218 (1997). 8. V. Cirimele, P. Kintz, C. Staub, and P. Mangin. Testing human hair for flunitrazepam and 7-aminoflunitrazepam by GC/MS-NCI. Forensic Sci. Int. 84:189-200 (1997). 9. V. Cirimele, P. Kintz, and P. Mangin. Determination of chronic flunitrazepam abuse by hair analysis using GC-MS-NCI. J. Anal. Toxicol. 20:596-598 (1996). 10. M. LeBeau, W. Andollo, W.L. Hearn, R. Baselt, E. Cone, B. Finkle, D. Fraser, A. Jenkins, J. Mayer, A. Negrusz, A. Poklis, H.C. Walls, L. Raymon, M. Robertson, and J. Saady. Recommendations for toxicological investigations of drug-facilitated sexual assault. J. Forensic Sci.

44:227-230 (1999).

11. V.I. Rickert, C.M. Wiemann, and A.B. Berenson. Prevalence, patterns, and correlates of voluntary flunitrazepam use. Pediatrics 103:1-5 (1999). 12. P. Kintz, A. Tracqui, and P. Mangin. Detection of drugs in human hair for clinical and forensic applications. Int. J. Leg. Med. 105: 1-4 (1992). 13. M. Rothe, F. Pragst, S. Thor, and J. Hunger. Effect of pigmentation on the drug deposition in hair of grey-haired subjects. Forensic Sci. Int. 84" 53-60 (1997).

Manuscript received March 16, 1999; revision received May 25, 1999.

435