Prevalence of Drugs Used in Cases of Alleged Sexual Assault

Transcription

1 Prevalence of Drugs Used in Cases of Alleged Sexual Assault M.A. ElSohly I and S.J. Salamone 2,' 1EISohly Laboratories, Inc., 5 Industrial Park Drive, Oxford, Mississippi and 2Roche Diagnostics, 1080 U.5. Highway 202, Somerville, New Jersey I Abstract In recent years, there has been an increase in the number of reports in the U.S. of the use of drugs, often in conjunction with alcohol, to commit sexual assault. A study was undertaken to assess the prevalence of drug use in sexual assault cases in which substances are suspected of being involved. Law enforcement agencies, emergency rooms, and rape crisis centers across the U.S. were offered the opportunity to submit urine samples collected from victims of alleged sexual assault, where drug use was suspected, for analysis of alcohol and drugs which may be associated with sexual assault. Each sample was tested by immunoassay for amphetamines, barbiturates, benzodiazepines, cocaine metabolite (benzoylecgonine), cannabinoids, methaqualone, opiates, phencyclidine and propoxyphene. The positive screen results were confirmed by gas chromatography--mass spectroscopy (GC-MS). In addition, each sample was tested for flunitrazepam metabolites and gamma-hydroxybutyrate (GHB) by GC-MS and for ethanol by gas chromatography-flame ionization detection (GC-FID). Over a 26-month period, 1179 samples were collected and analyzed from 49 states, Puerto Rico, and the District of Columbia. The states sending the most samples were California (183), Texas (119), Florida (61), Pennsylvania (61), New York (61), Minnesota (50), Illinois (47), Indiana (44), Michigan (40), Maryland (37), Virginia (32), and Massachusetts (31). Four-hundred sixty eight of the samples were found negative for all the substances tested; 451 were positive for ethanol, 218 for cannabinoids, 97 for benzoylecgonine, 97 for benzodiazepines, 51 for amphetamines, 48 for GHB, 25 for opiates, 17 for propoxyphene, and 12 for barbiturates. There were no samples identified as positive for phencydidine or methaqualone. In addition, 35% of the drug-positive samples contained multiple drugs. This study indicates that, with respect to alleged sexual assault cases, the prevalence of ethanol is very high, followed by cannabinoids, cocaine, benzodiazepines, amphetamines, and GHB. Although only a couple of substances have been implicated with sexual assault, this study has shown that almost 20 different substances have been associated with this crime. This study also raises the concern of illicit and licit drug use in sexual assault cases and suggests the need to test for a range of drugs in these cases. It also highlights the need to test for GHB, which is not generally tested for in a normal toxicology screen. Introdudion In recent years there has been an increase in the number of reports of the use of a number of drugs, often in conjunction *Author to whom correspondence should be addressed. salvatore.salamone@roche.com. with alcohol, to commit sexual assault (1-9). The reports have shown or suggested that these drugs are added to a victim's drink to produce an effect that leaves the victim in a semiconscious or unconscious state. Several drugs have been frequently mentioned as possibly being involved including flunitrazepam (Rohypnol) (2-4) and gamma-hydroxybutyrate (GHB). Flunitrazepam is one of a widely used class of prescription medications known as benzodiazepines. This drug is sold in 80 countries throughout the world but is not marketed in the United States or Canada. Flunitrazepam (10) is a low-dose, rapid-clearance benzodiazepine that is similar to alprazolam (11), triazolam (12), lorazepam (13), nitrazepam (14), bromazepam (15), and clonazepam (16). GHB is a substance that has recently gained attention in the toxicology and medical communities. This substance is known for its potency and can be easily prepared from readily available starting materials (2,5,6). Methods for the detection of many substances in urine are well defined and can be performed in most toxicology laboratories. Recently, several reports have described excretion patterns of low doses of flunitrazepam and analytical techniques that can be used to screen and confirm for the presence of low quantities of the drug and its metabolites. In cases where a 1-mg or 2-rag dose of flunitrazepam is consumed, the drug can be detected in the urine for up to 72 h by gas chromatography-mass spectroscopy (GC-MS) (17,18). As early as 1994, law enforcement and the rape-crisis-center community had identified a need for more sophisticated testing of suspected rape victims to determine if drugs or other substances may have been involved in sexual assault. As a result, a nationwide urine testing program was developed to assess the incidence of drugs in sexual assault and provide information for use in investigation of these crimes. Law enforcement agencies, emergency rooms, and rape crisis centers were offered the opportunity to submit urine samples that were collected from suspected victims of drug-induced sexual assault. Most of the urine samples were taken from these individuals within 72 h of the reported incident. Analysis was carried out by immunoassays for nine drug classes (amphetamines, barbiturates, benzodiazepines, cannabinoids, cocaine, methaqualone, opiates, phencyclidine, and propoxyphene) and all positives were confirmed by GC-MS. Regardless of the immunoassay result for benzodiazepines, each spec- Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission. 141

2 Journal of Analytical Toxicology, Vol. 23, May~June 1999 imen was analyzed for flunitrazepam metabolites by a highly sensitive GC-MS procedure. In addition, all specimens were analyzed for ethanol by GC-flame ionization detection (FID) and for GHB by GC-MS. The samples were collected over a 26- month period (between May 1996 and June 1998), and the results of this study are presented in this report. Methods Collection of samples Law enforcement officials, emergency room personnel, and rape crisis centers throughout the U.S. were offered the opportunity to send urine samples from individuals claiming to have been sexually assaulted when drug use was suspected. These officials were asked to submit samples under standard chainof-custody procedures to ElSohly Laboratories (Oxford, MS) for drug analysis. These samples were collected as part of the sexual assault investigation and were collected within 72 h (Figure 1) after the alleged incident occurred. The samples were refrigerated before testing and were tested within one week of collection. After testing, the samples were stored at -20~ Analysis Each sample was analyzed in the following manner: a sensitive GC-MS assay specific for flunitrazepam and its metabolites was run and had a limit of detection (LOD) of I ng/ml (17). GHB was analyzed by GC-MS (19), and ethanol was analyzed by GC-FID (19). An OnLine immunoassay (Roche Diagnostic Systems, Somerville, N J) screen for drugs of abuse was performed for each of the following nine drug classes, at the cutoff levels listed, on the COBAS MIRA analyzer (Roche Diagnostics): amphetamines 1000 ng/ml, barbiturates 200 ng/ml, benzodiazepines 50 ng/ml, cocaine metabolite 300 ng/ml, cannabinoids 50 ng/ml, methaqualone 300 ng/ml, opiates 300 ng/ml, phencyclidine 25 ng/ml, and propoxyphene 300 ng/ml (20). The immunoassays were run according to the manufacturer's recommended procedure with the exception of benzo- diazepines. The benzodiazepine assay was modified according to a previously published procedure that uses a highly sensitive 50-ng/rnL cutoff and ~-glucuronidase pretreatment to increase the assay clinical sensitivity (18). If the sample was at or above the immunoassay cutoff for a particular drug class a GC-MS confirmation assay was run to identify the specific drug or drugs (19). The confirmation assay for barbiturates was directed to identify butalbital, secobarbital, pentobarbital, butabarbital, and phenobarbital. The benzodiazepine confirmation assay was based on acid hydrolysis of the urinary metabolites and identification of the benzophenones corresponding to norfludiazepam, bromazepam, nitrazepam, temazepam, alprazolam, a-oh-alprazolam, triazolam, a-ohtriazolarn, diazepam, oxazepam, and lorazepam (19). The benzodiazepine GC-MS assay had an LOD of 5 ng/ml for all the benzodiazepines with the exception of alprazolam, a-hydroxyalprazolam, triazolam, and a-hydroxytriazolam. For these triazolobenzodiazepines, the LOD was 50 ng/ml. The confirmation test for opiates was directed to detect morphine and codeine, and the confirmation test for amphetamines was directed to identify amphetamine and methamphetamine (19). All other confirmation tests were able to identify the parent drug or the major metabolites of the parent drug (propoxyphene and norpropoxyphene, benzoylecgonine, 9- carboxytetrahydrocannabinol, and hydroxylated metabolites of methaqualone). Results Of the 1179 samples analyzed (Figure 1), 468 (39.7%) were negative for all of the substances tested and 711 (60.3%) samples were positive for one or more of the substances. Figure 2 shows the number of positive samples versus the total number of samples collected from each state. The most common substance found in the positive samples was alcohol with 451 samples, representing 40.8% of the samples, testing positive. The next most prevalent substance to be identified was cannabinoids (marijuana) with 218 positive cases, representing 18.5% of the samples. Benzodiazepines were identified Table I. Distribution of Positive Samples Containing Multiple Substances Also tested positive for Sample testing positive for N Alcohol Amphetamines Barbiturates Benzodiazepines Cocaine GHB Marijuana Opiates Propoxyphene Alcohot Amphetamines Barbiturates Benzodiazepines Cocaine Metabolite GHB Cannabinoids Opiates Propoxyphene

3 in 97 samples and represented 8.2% of the total number collected. Cocaine was found in 97 samples (8.2%), and amphetamines were identified in 51 samples (4.3%). GHB was found in 48 samples (4.1%), opiates in 25 samples (2.1%), propoxyphene in 17 samples (1.4%), and barbiturates in 12 samples (1.0%). No samples tested positive for phencyclidine or methaqualone. A high percentage of samples tested positive for several drugs, indicating multiple drug use as shown in Table I. There were 460 (39%) samples positive for one drug class, 251 (21.3%) samples positive for two drug classes, 153 (13%) samples positive for three drug classes, and 98 (8.3%) samples positive for four or more drug classes. In general, the most prevalent drugs found in alcohol-positive specimens were marijuana followed by cocaine and benzodiazepines. With respect to all of the other drug positive samples, the most prevalent substance was alcohol. The four most commonly identified substances showed the following relationship with respect to other drug use: of the 451 samples positive for alcohol, 12 were positive for amphetamines, 4 for barbiturates, 37 for benzodiazepines, 49 for cocaine, 16 for GHB, 75 for cannabinoids, 10 for opiates, and 8 for propoxyphene. Of the 218 samples positive for marijuana, 75 contained alcohol, 22 contained amphetamines, 6 contained barbiturates, 33 contained benzodiazepines, 35 contained cocaine, 10 contained GHB, 7 contained opiates, and 4 contained propoxyphene. With the 97 samples positive for cocaine, 49 contained alcohol, 6 contained amphetamines, 2 contained barbiturates, 17 contained benzodiazepines, 4 contained GHB, 35 contained marijuana, 9 contained opiates, and I contained propoxyphene. Of the 97 samples positive for benzodiazepines, 37 contained alcohol, 12 contained amphetamines, 7 contained barbiturates, 17 contained cocaine, 10 contained GHB, 33 contained marijuana, 9 contained opiates, and 6 contained propoxyphene. In addition there were 53 samples that were positive for three or more drug classes within the same sample. The distribution of all the samples with multiple substances detected can be seen in Table I. There were six different types of benzodiazepines identified by GC-MS in the benzodiazepine positive samples set. Of the 97 confirmed positive samples, 72 were identified as diazepam related samples containing diazepam, oxazepam, or nordiazepam. Four were positive for N-methyloxazepam (temazepam). Twelve of the benzodiazepine samples contained lorazepam, six contained flunitrazepam metabolites, eight contained Drug Total ~ ~'~ o.~ (~ 200- Z Is0-11x)- 50- o- alprazolam and its metabolites, four contained norfludiazepam (dalmane metabolites), and one contained bromoazepam. Some of the samples contained a mixture of two or more benzodiazepines. In addition to the 97 confirmed positives, there were an additional 50 samples identified positive by the screening assay that were negative for the specific benzodiazepines that the GC-MS confirmation test was directed to identify. A total of 49 states and the District of Colombia and Puerto Rico submitted samples for this study. The state of California submitted the most samples with 183, followed by Texas with 119, Florida with 61, Pennsylvania with 61, New York with 61, Minnesota with 50, Illinois with 47, Indiana with 44, Michigan E - :I Drug Figure 1. Distribution of positive test results on 1179 samples. Table II. Distribution of Specific-Drug-Positive Samples Versus Time After Alleged Incident Collected Alcohol 242 AmPhetamines 16 Barbiturates 1 Benzodiazepines 28 Cocaine Metabolite 39 GHB 16 Cannabinoids 56 Opiates 10 Propoxyphene 6 Time (h) > 72 Unknown I

4 with 40, Maryland with 37, Virginia with 32, and Massachusetts with 31. The remaining states sent in samples ranging in number from 1 to 27. The distribution of samples submitted from each state is summarized in Figure 2. Alcohol was the most prevalent substance found in the 10 states submitting the most samples with a percent positive ranging from 36% (Indiana) to 48% (Minnesota). The next most prevalent substance found in these 10 states was cannabinoids with a percent positive ranging from 15% (Pennsylvania) to 36% (Indiana). The distribution of benzodiazepines ranged from 5% (Michigan) to 14% (Indiana), and cocaine ranged from 6% (New York) to 16% (Texas). California had the highest percentage of positive specimens for amphetamines with 12.5%, and the other nine states had percent positives ranging from 0% to 7%. California also showed the highest positive rate of GHB with 8%, whereas the other states had percent positives ranging from 0% to 6%. Finally, Florida submitted five samples that were positive for propoxyphene, which represented 8% of the samples submitted from this state, and was the only state of the 10 that had more than one positive propoxyphene sample. The majority (65.3%) of the 1179 samples were collected within 24 h of the estimated time of the assault, and 96% of the samples were collected within 72 h (Figure 3). The majority of samples identified as positive were collected within the first 24 h (Figure 4). Two-hundred ninety eight samples (41.9%) were collected within 12 h of the incident and 264 (37.1%) were collected between 12 and 24 h. There were 43 positive samples (6.1%) collected between 25 and 36 h, 49 positive samples (6.9%) collected between 37 and 48 h, and 57 (8%) positive samples collected at times greater than 48 h. The distribution of specific-substance-positive samples (Table II) from the time of collection paralleled that of the distribution of total samples collected for each time interval for amphetamine, benzodiazepines, cocaine, GHB, and marijuana. The total percentage of samples collected in the first 24 h was 68.3%, and the percentage of samples positive for amphe-tamine, benzodiazepines, cocaine, GHB, and marijuana ranged between 66 and 73%. The specific drugs that did not parallel the total distribution were alcohol with 90.5%, opiates with 80.0%, and propoxyphene with 76.5% positives collected within 24 h of the incident. This indicates that the positive rate for these drugs is representative for the entire set of samples, including those samples collected at 48 and 72 h. Conversely, if the percentage of positive samples for a particular substance, as in the case of alcohol, is higher than the percentage of total samples collected in that time interval then the prevalence of this substance is probably understated for the entire sample set. This difference could be explained by the short urinary tl~ for alcohol that would make the detection of alcohol beyond 24 h difficult. With respect to the 12 positive barbiturate specimens, 41% of these samples were collected within 24 h after the incident. 0/2 2/ /13 of Co~umb~: 0/I US Gov~nma1:s't2 011 sq 8/t0 United States (AK & HI ~set) by % posiliv e 9 80 to 100 (5) []60to 8o (19) 1740to 6o (is) []20to 40 (6) [] 0 to 20 (5) Puerto Rico: I/2 Figure 2. Number of positive samples over the total number of samples collected by each state. 144

5 Conclusions A nationwide study was undertaken to determine the prevalence of different drugs in alleged sexual assault cases by analysis of urine specimens collected from victims by law enforcement officers, emergency room personnel, or rape crisis centers. Most of the specimens were collected within 24 h of the incident, and more than 90% of the specimens were collected within 72 h. Between May 1996 and June 1998, 1179 cases were investigated and 39.0% of the urine specimens collected were found to contain at least one drug with 42.7% of the positive specimens containing more than one drug. The drug found in the majority of the positive specimens was alcohol. The next most prevalent drug found in the positive specimens was cannabinoids (marijuana), followed by the general class of benzodiazepines, cocaine, amphetamines, and GHB. The pattern of prevalence of the different drugs differed among individual states. In particular, California showed the highest percentage of positive samples for amphetamines and GHB, Texas for cocaine, Indiana for benzodiazepines and Florida showed the highest positive rate for propoxyphene. Because of the lack of correlation of drug concentrations in urine with impairment of the individual, interpretation of this data should be made with caution. The presence of these substances in the urine of individuals does not necessarily indicate the involvement "~ r of these drugs in the sexual incident and may ~" reflect an exposure to the drug before or after the incident. Diazepam, oxazepam, nordiazepam, and their related benzodiazepines made up 72 of the 97 positives. The parent drugs of these com- pounds are frequently prescribed for many reasons, therefore it is not surprising that they appear in this type of analysis. Because a single dose of diazepam can be detected in the urine for longer than two weeks after ingestion (21), the prevalence of this benzodiazepine class in a high number of cases may reflect ingestion of the drug before the day of the incident and not the use of it during the incident. The prevalence of the lower dose high clearance benzodi- azepines was considerably lower. Alprazolam > was found in 8 samples, flunitrazepam in 6 samples, and lorazepam in 12 samples. The testing procedure selected for alprazolam was d the least sensitive (GC-MS at 50 ng/ml), which =, may have understated the prevalence of the drug in this study. The sensitivity of the GC-MS procedure used for lorazepam and flunitrazepam was higher. Although the testing procedure selected for flunitrazepam was the most sensitive (GC-MS at I ng/ml), only six specimens were found to contain the drug or its metabolites. The six positive samples were collected within a 24-h period of the incident, and the concentrations found in the urine reflect that the individuals had ingested a low to moderate dose of the drug (2-4 rag) (17,18). The 50 samples that screened positive for benzodiazepines by immunoassay and did not confirm by GC-MS may have been benzodiazepines that the GC-MS was not directed to detect or may have been due to the lower sensitivity of the GC-MS procedure used to confirm the triazolobenzodiazepines. In order to better elucidate these unconfirmed benzodiazepines, a more sensitive procedure for the triazolobenzodiazepines would be appropriate. The prevalence of GHB is of a special concern because there are no screening tests available to detect this drug. GHB is a potent tranquilizer that is cleared from the body rapidly (22-24). Because laboratories do not routinely test for this drug, there is a potential for undetected use of GHB in sexual assault. The results of this study demonstrate that a wide range of drugs could be involved in cases of sexual assault. With P-eg Time (h) Figure 3. Time interval from alleged drug ingestion to collection of sample. 20(]" IlXl- Q~ ~ ~ Tlme postlncldent (h) Figure 4. Distribution of any positive sample by time of collection from incident. m Unknown 145

6 current technologies most of these drugs can be detected within a 24-h period of the incident. Rape kits should contain a urine collection container and urine samples should be collected from a sexual assault victim as soon as possible from the time of the incident. This study also demonstrated that no one drug was associated with sexual assault, rather many drugs appear to be associated with this crime. Furthermore, attention should be given to other lesser known drugs such as GHB that should be included in the battery of tests used in sexual assault cases. References I. J. Rhynard, M. Krebs, and J. Glover. Sexual assault in dating relationships. J. Sch. Health 67:89-93 (I 997). 2. C. Bismuth, S. Dally, and S.W. Borron. Chemical submission: GHB, benzodiazepines, and other knock out drops. J. Toxicol. Clin. Toxicol. 35" (1997). 3. S.R. Calhoun, D.R.Wesson, G.P. Galloway, and D.E. Smith. Abuse of flunitrazepam (Rohypnol) and other benzodiazepines in Austin and south Texas. J. Psychoactive Drugs 28: (1996). 4. D. Anglin, K.L. Spears, and H.R. Hutson. Flunitrazepam and its involvement in date or acquaintance rape. Acad. Emerg. Med. 4" (1997). 5. H.K. Louagie, A.G. Verstraete, S.C. De, D.G. Baetens and P.A. Calle. A sudden awakening from a near coma after combined intake of gamma-hydroxybutyric acid (GHB) and ethanol. J. Toxicol. Clin. Toxicol. 35: (1997). 6. From the CDC: gamma hydroxy butyrate use---new York and Texas, J. Am. Med. Assoc. 227:1511 (1997). 7. D. Brahams. Benzodiazepines and sexual assault, Canada [news]. Lancet 337: (1991 ). 8. D.R. Wesson and W. Ling. Addiction medicine. ]. Am~ Med. Assoc. 275: (1996). 9. S. Ranieri-Collins. Alcohol-related sexual assualt. Trends Health Care Law Ethics 10(3)" (1995). 10. H.G. Boxenbaum, H.N. Posmanter, and T. Macasieb. Pharmacokinetics of flunitrazepam following single- and multiple-dose oral administration to healthy human subjects. J. Pharmacokinet. Biopharm. 6: (1978). 11. G.W. Dawson, S.G. Jue, and R.N. Brogden. Alprazolam: a review of its pharmacodynamic properties and efficacy in the treatment of anx- iety and depression. Drugs 27: (1984). 12. G. Baktir, H.U. Fisch, P. Huguenin, and J. Bircher. Triazolam concentration-effect relationships in healthy subjects. Clin. PharmacoL Ther. 34" (1983). 13. D.J. Greenblatt, R.T. Schillings, and A.A. Kyriakopoulos. Clinical pharmacokinetics of Iorazepam. I. Absorption and disposition of oral 14C-lorazepam. Olin. PharmacoL Ther. 20" (1976). 14. J. Rieder and G. Wendt. Pharmacokinetics and metabolism of the hypnotic nitrazepam. In The 8enzodiazepines, S. Garattini, E. Mussini, and L.O. Randall, Eds. Raven Press, New York, NY, 1973, pp S.A. Kaplan, M.L. Jack, R.E. Weinfeld, W. Glover, L. Weissman, and S. Cotler. Biopharmaceutical and clinical pharmacokinetic profile of bromazepam. ]. Pharmacok inet. 8 iopharm. 4:1-16 (1976). 16. A. Berlin and H. Dahlstrom. Pharmacokinetics of the anticonvulsant drug clonazepam evaluated from single oral and intravenous doses and by repeated oral administration. Eur. J. Clin. Pharmacol. 9: (1975). 17. 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: (1997). 18. S.J. Salamone, S. Honasoge, C. Brenner, 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: (1997). 19. Standard Operating Procedure, EISohly Laboratories, Oxford, MS, Roche Diagnostics, Somerville, NJ, package insert, OnLine Assays for Amphetamines, Barbiturates, Benzodiazepines, Benzoylecgonine, Cannabinoids, Methaqualone, Opiates, Phencyclidine and Propoxyphene. Sept S.A. Kaplan, M.L. Jack, K. Alexander, and R.E. Weinfeld. Pharmacokinetic profile of diazepam in man following single intravenous and oral and chronic oral administrations. J. Pharm. Sci. 62: (1973). 22. G. Tunnicliff. Sites of action of gamma-hydroxybutyrate (GHB)--a neuroactive drug with abuse potential. J. Toxicol. Olin. Toxicol. 35: (1997). 23. C. Marwick. Coma-inducing drug GHB may be reclassified [news]. J. Am. Med. Assoc. 277" (1997). 24. R.C. Baselt and R.H. Cravey. Disposition of Toxic Drugs and Chemicals in Man, 4th ed. Chemical Toxicology Institute, Foster City, CA, 1995, pp Manuscript received February 4, 1999; revision received March 30,