Temporal Indication of Marijuana Use Can Be Estimated From Plasma and Urine Concentrations of Ag-Tetrahydrocannabinol, 9

Transcription

1 Temporal Indication of Marijuana Use Can Be Estimated From Plasma and Urine Concentrations of Ag-Tetrahydrocannabinol, 11-Hydroxy-A 9 -Tetrahydrocannabmol, and 11.Nor-Ag-Tetrahydrocannabinol-9-Carboxylic Acid*" Joseph E. Manno 1,2,*, Barbara R. Manno 2, Philip M. Kemp 3, Dempsey D. Alford 1, Imad K. Abukhalaf 2,w Mary E. McWilliams 4,~, Frances N. Hagaman 2, and Mary Jo Fitzgerald 2 Departments of Emergency Medicine (Section of Toxicology) z, Psychiatry 2, and Neurology 4, Louisiana State University Health Sciences Center, Shreveport, Louisiana and 3Chief Medical Examiner's Office, Oklahoma City, Oklahoma Abstract I Current technology establishes marijuana use based upon detection of the pharmacologically inactive cannabinoid metabolite (11-nor- Ag-carboxy-tetrahydrocannabinol-9-carboxylic acid, THC-COOH) in urine. No accurate prediction of time of use is possible because THC-COOH has a half-life of 6 days. To determine if a temporal relationship between marijuana use and metabolite excretion patterns could be established, eight healthy user-volunteers (18-35 years old) smoked marijuana cigarettes containing 0% (placebo), 1.77%, and 3.58% Ag-tetrahydrocannabinol (THC). Plasma and urine were collected prior to smoking, 5 rain after smoking, and hourly thereafter for 8 h for measurement of cannabinoid concentrations by gas chromatography-mass spectrometry. Mathematical models proposed for determination of recent marijuana use were applied to data from this study and verified the temporal use of marijuana. One subject, who later admitted chronic marijuana use (urine baseline THCCOOH, ng/ml; plasma, 75.5 ng/ml), excreted 8]3- dibydroxy-thc, peaking 2 b postsmoking (92.3 ng/ml). Urinary THC, the psychoactive component of marijuana, concentrations peaked 2 h after smoking and declined to assay limit of detection (LOD) (1.5 ng/ml) by 6 h. 11 -Hydroxy-A%tetrahydrocannabinol (11-OH-THC) and THCCOOH were detectable for the entire 8-h testing period but continued to decrease. Urinary concentrations of THC greater than 9 Presented in part at the t 995 Society of Forensic Toxicologists meeting, Baltimore, MD and at the 1998 California Association of Toxicologists meeting, San Francisco, CA. ' Supported in part by NIDAgrant no. DA * Author to whom correspondence should be addressed: }oseph E. Manno, Ph,D,, LSU Health Sciences Center, Department of Emergency of Medicine (Section of Toxicology), P.O. Box 33932, Shreveport, LA Current address: Morehouse School of Medicine, Department of Pharmacology and Toxicology, Atlanta, GA. Current address: Private practice (neurology), Bossier City, LA. 1.5 ng/ml suggests marijuana use during the previous 8-h time period. Introduction Marijuana has been used for medicinal and recreational purposes for thousands of years (1,2) and persists as the most widely used illicit drug in the United States (3). In addition, the DAWN Mid-Year 2000 Preliminary Emergency Department Data indicates that the number of marijuana/hashish Emergency Departments mentions increased fourfold, from < 10,000 in January-June 1990 to nearly 50,000 in January-June 2000 (4) and is the most commonly detected drug of abuse in workplace urine drug screens (5). Smoking is the preferred route of administration because of the rapid absorption and distribution of the primary psychoactive component, A~ (THC), to brain (6,7). It has been found that THC could be detected in the plasma after the first puff of a marijuana cigarette (7). Plasma concentrations peaked during the smoking process, and dropped to < 10 ng/ml within 1 h (8,9). This rapid reduction in plasma concentrations has been attributed to the highly lipophilic nature of THC resulting in a rapid distribution from blood to the tissues, and to its rapid conversion to more polar (hydrophilic) metabolites (10). The metabolism of THC in humans has been extensively reviewed (11,12). Metabolite composition varies with specimen source (animal, human) and experimental conditions (13). In humans, in vitro hepatocyte studies have demonstrated that initially THC is hydroxylated via a cytochrome P450 enzyme subfamily (14,15) primarily at the C-11 position, forming the active 538 Reproduction (photocopying) of editorial content of this journal is prohibited without pub]isher's permission.

2 8-Hydroxy-A9-THC A9-THC / x, 11-OH-A9-THC 8, 11-Dihydroxy- ~ 11-Nor-A9-THC - Ag-THC 9-carboxylic acid \ / Glucuronide Conjugates All are excreted in urine Figure 1. Scheme for biotransformation of THC. Table I. Huestis et al. (21) Mathematical Models for Predicting Time of Marijuana Use Model I LogT= ( x Log [THC}) Model II LogT= (0.576 x Log [THCCOOH +THC]) metabolite, 11-hydroxy-A9-tetrahydrocannabinol (11-OH-THC). Following smoking, this metabolite has been reported to be less than 10% of THC concentration in plasma (16). Oxidation of 11- OH-THC results in the formation of the most abundant THC metabolite (Figure 1) found in plasma and urine, 11-nor-Ag-carboxy4etrahydrocannabinol-9-carboxylic acid (THC-COOH) (17). THC-COOH concentration in plasma rises slowly but eventually increases above THC concentration as the latter quickly falls (7,10). The short plasma half-life of THC requires that blood must be collected within 2 h of use in order to have a detectable quantity of the cannabinoid present. Under the ideal circumstance of a rapid collection time, an individual would have to be actively smoking marijuana at the time of an adverse event in order to be able to measure meaningful concentrations of THC in plasma. Consequently, because ideal situations do not always exist in cases requiring drug testing, clinical and forensic toxicologists are often called upon to interpret plasma and/or urine concentrations of THC-COOH. Because THC-COOH is not psychoactive, no relationship between plasma and/or urine concentrations and human psychomotor performance can be made. The long halflife of THC-COOH and normal renal physiology also precludes establishing an accurate relationship between time or marijuana use and plasma and/or urine concentrations. In fact, as stated, this temporal relationship has not been easy to establish with THC because of its rapid redistribution from plasma, lipid solubility and to its complex metabolism. In recent years, research has focused upon drug metabolites and metabolic patterns of THC in biological matrices and any temporal relationship these compounds may have with physiological and psychological effects. Using drug metabolite/parent ratios for determining time since use has been suggested (18). A dihydroxy metabolite (813,11-dihydroxy-THC) that is eliminated within 24 h in urine has also been suggested as a possible marker for recent marijuana use (19). Additionally, we have previously suggested a possible relationship between performance decre- ments and plasma cannabinoid concentrations up to 2 h after smoking (20). Using data from published reports of THC/metabolite blood levels and from their own research mathematical models (Table I) have been developed based upon plasma THC levels and THC-COOHffHC ratios for predicting the time that a person smoked marijuana (21). More recently, Fraser and Worth (22) applied both the Huestis and Cone formula (Model II) and the Manno et ai urine THC-COOH/creatinine formula to their own research findings to predict new use of marijuana. They have recommended that the Huestis and Cone (21) ratio is best used in clinical settings because of its lower false-negative rate (24%). On the other hand, the Manno et al. (23,24) ratio was recommended for use in legal situations because of its lower false-positive rate of 0.1% as reported by Huestis and Cone (25). Previous work in this laboratory has shown for the first time that both THC and ll-oh-thc are present in urine as glucuronide conjugates, and are easily detectable after enzymatic hydrolysis (26,27). This has been further elucidated and confirmed by EISohly and Feng (28) in their work on cannabinoid metabolites in meconium. The aim for the present study was twofold. First, confirmation of the presence of 89,11-dihydroxy-THC, the dihydroxy metabolite reported by McBurney et al. (19), and determination of its utility as a marker in urine to estimate time of marijuana use. Second, to establish quantitative measures of THC and 11-OH- THC, both pharmacologically active cannabinoids, in plasma and urine after smoking marijuana in a controlled study, in order to establish their value as markers for estimation of time of marijuana use. Methods Subjects Eight subjects (4 male and 4 female; ages years) with a self-reported history of light marijuana use (1-3 cigarettes per week or less) were recruited through an advertisement in a local newspaper. Prior to acceptance in the study, the subjects underwent extensive physical and psychiatric examinations to determine that they were in good health. The physical examination included a blood chemistry panel, complete blood count, urine drug testing, electrocardiogram (ECG) and pregnancy testing for females. The blood chemistry panel consisted of sodium, potassium, chloride, total CO2, anion gap, urea nitrogen, creatinine, blood urea nitrogen (BUN)/creatinine ratio, glucose, calcium, phosphorus, uric acid, triglycerides, cholesterol, total protein, albumin, total bilirubin, alkaline phosphatase GOT (AST), and total LDH tests. Urine drug testing was done by immunoassay (EMIT M) for amphetamines, barbiturates, benzodiazepines, cocaine metabolite, opiates, phencyclidine, proproxyphene, ethanol and THC. The psychiatric examination consisted of an interview by a psychiatrist and evaluation using the Schedule for Affective Disorders and Schizophrenia-Life Time Version (SADS-L)(29). Following final acceptance on the study, the subjects were tested at the beginning of each experimental session for breath 539

3 alcohol and drugs with urine immunoassay (Triage TM, Biosite Diagnostics, Inc., San Diego, CA) for all of the above drug classes except propoxyphene. Female subjects were tested at each session for pregnancy [(Icon II HCG (urine), Hybritech, Inc., San Diego, CA]. Testing sessions were spaced one week apart to allow for significant clearance of the previous dose of marijuana. Subjects were instructed not to use marijuana or other drugs for the duration of the study. Upon arrival at the laboratory on test days, the subject provided a urine sample for drug and pregnancy testing (if applicable). In addition, they were tested for breath alcohol (Alcosensor III, Intoximeter, Inc., St. Louis, MO). Tobacco use was restricted on testing days. All subjects provided informed consent and all were compensated for their time in the laboratory. All protocols and informed consent forms were approved by the LSU Health Sciences Institutional Review Board for Human Research. Sample collection Upon arrival at the laboratory, an 18 gauge x ll/4-in. JELCO TM winged i.v. catheter (Critikon, Tampa, FL) equipped with an Interlink TM Injection Site (Baxter Healthcare Corp., Deerfield, IL) was inserted into a vein in the forearm. Blood samples (5-7 rnl) were collected using 7-mL Vacutainer tubes containing 100 USP units of lithium-heparin (Becton Dickinson, Rutherford, N J) prior to smoking, 5 min after smoking and hourly for 8 h after smoking. The catheter was flushed with sterile saline after each blood draw. The tubes were immediately centrifuged, and the plasma was transferred to silanized ram glass, screw-capped culture tubes and stored frozen (-20~ until time of analysis. At the time of analysis, the samples were allowed to warm to room temperature. If the samples were not visually clear, they were centrifuged prior to analysis. Drug administration Marijuana cigarettes were obtained from Research Triangle Institute (Research Triangle Park, NC) with the approval of the National Institute on Drug Abuse (NIDA). A single marijuana cigarette (0% THC [placebo], 1.77% THC or 3.58% THC) was smoked each session that were scheduled at weekly intervals. Subjects were cued during the paced smoking process utilizing computer generated instructions for standardization of the smoking procedure for inhalation (3 s), breath holding (20 s) and a relaxation period (34 s). The timing was similar with that previously reported by Huestis et al. (7). Eight puffs were administered over a 7.6-rain period of time. The cigarettes weighed an average of 730 mg (placebo), 857 mg (1.77% THC), and 753 mg (3.58% THC) for calculated doses of 0,15.2, and 26.9 mg of THC, respectively. Subjects were dosed according to randomized, double-blind, Latin-square, crossover design at weekly intervals on an outpatient basis. Analysis of cannabinoids by gas chromatography-mass spectrometry (GC-MS) The cannabinoids were separated from the biologic matrices utilizing a procedure previously described by Kemp et al. (26). Briefly, plasma and urine samples were hydrolyzed with bacterial ~-glucuronidase (Escherichia coli). After extraction with hexane/ethyl acetate (7:1), analysis and quantitation were performed by GC-MS. A Hewlett-Packard (HP) GC-MS consisted of an HP 5890 GC with an HP-5 MS capillary column, an injector with electronic pressure programming, and an HP 5972 mass selective detector operated in the electron impact ionization (EI) mode. The cannabinoids were detected, using selected ion monitoring (SIM), as their trimethylsilyl derivative that enhanced their chromatographic separation and mass spectral characteristics. Statistical analysis Data were examined by with analysis of variance (ANOVA) for statistical differences. Significant differences were further identified using the Neumann-Keuls post-hoc test. Areas under the curve (AUC) were calculated using the tapezoidal rule method. Regression analysis was done using the computer program "Pharmacologic Calculation System" (30). Results and Discussion Cannabinoids in plasma Plasma concentrations of the cannabinoids following the smoking of a single marijuana cigarette are shown in Tables II and III (1.77% and 3.58% THC, respectively). Pharmacodynamic comparisons for mean plasma data are summarized in Table IV. The mean THC concentrations for the 1.77% and 3.58% THC cigarettes over the 8-h time course are represented graphically in Figure 2. THC concentrations peaked during the 5-rain interval after smoking (Figure 2). Mean peak concentrations of THC were higher with 3.58% than the 1.77% dose (p < 0.05). Peak plasma concentrations ranged from 1.5 to 30.4 ng/ml with a mean ( SEM) of ng/ml for the 1.77% dose and from 17.2 to 63.0 ng/ml with a mean of ng/ml for the 3.58% marijuana cigarettes, establishing a dose-dependent increase in plasma concentration of THC. At 1 h after smoking, mean plasma concentrations had fallen to 5.5 and 2.5 ng/ml for the high and low dose, respectively, a rate of 30.7 and 11.9 ng/ml/h. At 2 h, plasma concentrations after the low dose had fallen below the detection limit of 1.5 ng/ml THC, but the high dose remained detectable up to 3 h after administration. The AUC for mean plasma THC was ng-h/ml and ng-h/ml for the 1.77% THC and 3.58% THC (Table IV), respectively, again demonstrating a dose-dependent plasma-thc relationship (p < 0.01). The rapid disappearance of THC from the plasma is in agreement with other studies that investigated plasma cannabinoid levels after smoking (10,19). Others have reported that THC levels fell below 5 ng/ml within 2 h after smoking 1.75% and 3.55% THC cigarettes (7). Rapid disappearance of THC has been attributed to the lipophilic nature of the compound resulting in the redistribution of THC to the tissues (31). In addition, THC is rapidly metabolized as evidenced by the presence of 11-OH-THC as early as the 5 min postsmoking time point in this study and even after the first puff in one subject in another study (7). A dose-dependent plasma concentration increase also occurred for the primary active metabolite, 11-OH-THC, which 540

4 Table II. Plasma Cannabinoids (ng/mt)--1.77% THC Marijuana Cigarette--Time Post-Smoking (min) Subjed THC 11.OH-THC THGCOOH THC 11.OH-THC THC.COOH THC 11.OH.THC THC-COOH , * Mean t SEM THC 11-OH-THC THC-COOH THC 11.OH.THC THC-COOH THC 11.OH.THC THC.COOH , ,2 0 1, , Mean SEM 0, N THC 11-OH.THC THC.COOH THC 11-OH-THC THC-COOH THC 11-OH-THC THC-COOH Mean SEM N THC 11-OH-THC THGCOOH I , ,7 Mean SEM N * No specimen available. + Values derived without inclusion of Subject S because of self-reported marijuana use within 24-h period prior to participation in low-dose session. Subject did not make disclosure until testing was completed for the day. * N= numberof specimens, 541

5 Journal of Analytical Toxicology, Vol. 25, October 2001 Table III. Plasma Cannabinoids (ng/ml)--3.58% THC Marijuana Cigarette--Time Post-Smoking (min) Subject THC 11.0H.THC THC-COOH THC 11-OH.THC THC.COOH THC 11-OH.THC THC-COOH , , , , , ,3 19,2 Mean 0 0,2 16,3 33, , EM , N f THC 11-OH-THC THC-COOH THC 11-OH-THC THC-COOH THC 11-OH-THC THC-COOH 1 -* , , , , ,5 125, , , Mean , , SEM ,3 14,7 0, N THC 11.0H.THC THC-COOH THC 11-OH.THC THC.COOH THC 11-OH-THC THC.COOH , , , , t 0 0 t Mean SEM ,3 10,1 N THC 11.0H-THC THC.COOH , Mean ,8 SEM N * No specimen available. t N = number of specimens. 542

6 peaked at 5 rain after smoking (Figure 2). Peak concentration ranged from 0 ng/ml to 6.3 ng/ml with a mean (+ SEM) of ng/ml for the 1.77% THC cigarette and from 4.1 to 21.3 ng/ml with a mean of ng/ml, for the 3.58% THC cigarette (< 0.05) (Tables II and III). At I h post-smoking, the plasma concentrations were 4.2 and 1.2 ng/ml for the high and low doses, respectively. The clearance rate was calculated to be 4.5 and 1.5 ng/mi,/h, more than 85% slower than THC. AUCs (Table IV) were ng-h/ml and ng-h/ml for the low and high doses, respectively (p < 0.05). The slower disappearance of 11-OH-THC from plasma has also been observed in other studies, and may be associated with differences between the enzymes responsible for the oxidation of THC and 11-OH-THC (7,10). This hypothesis is supported by animal research by Gill and Jones (32), in which SKF525a was used to inhibit hepatic enzymes. The results showed that the 7o -f T T Plasma 60 THCCOOH 50 4o 3o 2o lo u~ 8 ' o - 35 Plasma 11-OH-THC Plasma THC ~ ~ 0~1.77%THC I 3.58%THC ' lo! Time after smoking (min) Figure 2. Plasma concentrations (mean SEM) of THC and its major metabolites after smoking marijuana. hydroxylation of THC to 11-OH-THC was only slightly inhibited while the metabolism of 11-OH-THC to subsequent metabolites demonstrated a greater inhibition. Later work using human liver in vitro has shown that the parent THC and metabolite 11-OH- THC were metabolized by different subfamilies of the cytochrome P450 enzyme system (15). At the peak time concentration point (5 rain) when both THC and 11-OH-THC were present, the concentrations of the 11-OH- THC were 7.8 to 123% of the THC plasma levels (mean = 30.7%). Other investigators have reported that, following marijuana smoking, peak 11-OH-THC concentrations are only 5 to 10% of THC concentrations (7,10). The higher percentage of 11-OH- THC obtained in the present work may be a result of the incorporation of a hydrolysis step using ~-glucuronidase from the bacteria E. coli (26). We have found that the bacterial enzyme is much more reactive toward the ether bonded glucuronides of the hydroxylated metabolites of THC in urine (27) than glucuronidase from other sources. McBurney et al. (19) used [3-glucuronidase from mollusk (Helixpomatia), which we found to be much less reactive toward 11-OH-THC-glucuronide (27). Huestis et al. (7) did not use a hydrolysis step in their extraction procedure. The results of the present study indicate that there may be a higher concentration of the 11-OH-THC glucuronide in plasma than previously reported. ElSohly and Feng (28) have more recently shown similar results in meconium with enzymatic hydrolysis as Kemp et al. (26,27) reported in plasma and urine. It is interesting that EISohly and Feng (28) found that cleanup steps, for example, back titration and increase of glucuronidase activity, was necessary in meconium to increase cannabinoid yield because of the presence of high background interference (e.g., lipids). It is obvious that additional work is needed in characterizing recovery of the THC metabolites from biologic fluids and tissues because 8~,11-diOH-THC was detected in the neutral fraction of the hydrolysis with Kemp and co-workers' method (26) and ElSohly and Feng found the metabolite in acidic fraction with their method (28). Because of our finding of increased ratios of plasma 11-OH- THC relative to THC, we attempted to determine if a similar increase in relative concentration of other hydroxylated metabolites would occur, thereby providing more information regarding metabolic patterns that would demonstrate a temporal relationship to marijuana usage patterns. Following intravenous administration, plasma levels of 8[3-OH-THC have been shown to be similar to those of 11-OH-THC, both being 5 to 10% of the THC concentration (10,16). Plasma concentration of 8o~-OH-THC were reported lower than the 8[Msomer. In addition, the 8,11- dihydroxylated species were present in plasma in concentration similar to the 8~-OH-THC. In the present work, no 8-OH or 8- dioh metabolites of THC were detected in plasma, the limits of detection for these compounds ranged from 0.8 to 1.1 ng/ml (26). The explanation for this is not apparent as the 11-OH-THC and 8~-OH-THC have been shown to follow similar patterns of metabolism and reach similar concentrations in plasma (16). Plasma concentrations of THCCOOH did not reach peak concentrations until I to 2 h after smoking (Figure 2). The plasma levels declined very gradually, but never returned to baseline values for the duration of the study (8 h). Pharmacokinetic parameters for THCCOOH after acute marijuana use were diffi- 543

7 cult to ascertain, presumably because of the often reported accumulation of this metabolite that occurs with repeated marijuana use (33,34). Peak plasma concentrations of THCCOOH after Table IV. Pharmacodynamic Comparisons by THC Dose and Matrix* Parameter Placebo P~'~ 1.77% THC p % THC p1-3 Plasma Peak, ng/ml THC 0.0 (8) t (8) (8) 11-OH-THC 0.0(8) (8) (8) THC-COOH (8) NS (8) NS (8) AUC (O-8 h) THC 0.0(8) (8) (8) 11-OH-THC 1.9 _+ 1.9 (8) NS 7.6 +_. 3.2 (8) (8) THC-COOH (8) NS (8) NS (8) Urine Peak, ng/ml THC 0.0 (5) NS (5) NS (7) 11-OH-THC (5) (5) NS (5) THC-COOH 59.4_ 51.1 (3) NS (4) NS (3) AUC (0-8 h) THC (8) (8) (8) 11-OH-THC (8) NS (8) NS (8) THC-COOH (8) NS (8) NS (8) * All data are mean -+ SEM; p1-2 = placebo vs. 1.77%; p2-3 = 1.77% vs. 3.58%; p1-3 = placebo vs. 3.58%. * Number of subjects in parentheses. Table V. Mean Predicted Exposure Times with Models I and I1' Predicted time (h) Time (h) t n Model 1 n Model II 0~ ND* ND * Huestis et al. (21). Time = hours from the end of smoking. * ND = No THC detected. Table VI. Accuracy of Models I and I1' in Predicting Time of Marijuana Exposure Time (h)* smoking the 3.58% THC cigarette ranged from 10.9 to ng/ml at the 2-h time point (mean = ng/ml). Similar values were obtained after smoking (7) and following intravenous administration of 5 mg THC (35). Clinical studies in a laboratory setting with controlled smoking are important for the characterization of the time course of detection of THC and its metabolites in biologic matrices. Studies like this project provide data for establishing any temporal relationship between cannabinoid concentrations and behavioral effects. A mathematical model for the prediction of the time of NS O.Ol O.Ol NS 95% CP Current Study Model I Model!1 n Model I n Model II * Huestis et al. (21). CI = Confidence intervals. * Time = hours from the end of smoking. marijuana use, based on plasma cannabinoid concentrations, has been reported (21). Linear regression analysis of plasma THC concentration versus time (in hours) provided Model I for predicting time since exposure of marijuana (Table I), where T = time (in hours) since exposure to marijuana. In addition, a second model was based on the plasma ratio [THCCOOH] - [THC] as shown in Table I. Using the plasma concentrations measured in our work, Model I was evaluated for its accuracy in predicting the time that the marijuana NS was smoked. Table V shows the times of marijuana exposure predicted from Model I from plasma THC concentration obtained after subjects smoked. The error in 0.01 prediction increased with longer time since exposure, o.o5 an effect previously reported (21). Predicted times of N5 exposure using Model I (THC concentration in plasma) were acceptable with mean percent error of % (mean _ SD). Accuracy of Model I was also assessed using the 95% confidence interval reported (21). The model is considered accurate if the predicted exposure time calculated from Model I fall within the 95% confidence intervals. Table VI shows how plasma THC concentrations from the present study fit into the 95% confidence limits of the Huestis model. The Model I predictions were overestimated up to the 2-h post-smoking point, then underestimated up at the 4-h post-smoking time point. The second proposed mathematical model, Model II, for predicting time of marijuana exposure (Table I) was tested for accuracy using the ratio of the plasma concentrations of THCCOOH to THC ([THCCOOH] :- [THC]). Table V compares the predicted exposure times using this model with the actual times after marijuana smoking from our study. The accuracy was acceptable until the 4.0-h time point (7.1 h predicted). This agreed with the report of Huestis et al. (21), which found a mean error of 4.22 h at the 4-8 h post-exposure times. Model II was also evaluated for accuracy using the 95% confidence limits reported for the model (21). Fit to the model is considered accurate if the predicted time 1.7 of exposure falls within the 95% confidence limits. The 2.6 actual times of exposure of 24 of 25 (96%) samples mea- 7.1 sured in this study fell within the 95% confidence limits shown in Table VI. In each case, Model II tended to overestimate the time of exposure. The two mathematical models (7) for determination of time since exposure were validated by the data in our 544

8 Table VII. Urine Cannabinoids Concentrations (ng/ml)--1.77% THC Marijuana Cigarette--Time Post-Smoking (min) Subject THC 11-OH.THC THC-COOH THC 11-OH-THC THC.COOH THC 11.OH-THC THC-COOH 1 -* Mean t SEM N* THC ll-oh-thc THC-COOH THC 11-OH-THC THC-COOH THC 11-OH-THC THC-COOH Mean SEM N THC 11-OH-THC THGCOOH THC ll.oh-thc THC-COOH THC 11-OH-THC THGCOOH I Mean SEM N THC 11-OH-THC THC-COOH I , , Mean SEM N * No specimen available. t Values derived without inclusion of Subject S because of self-reported marijuana use within 24-h period prior to participation in low-dose session. Subject did not make disclosure until testing was completed of the day. * N = number of specimens. 545

9 Table VIII. Urine Cannabinoids Concentrations (ng/ml)--3.58% THC Marijuana Cigarette--Time Post-Smoking (rain) Subject THC 11-OH-THC THC-COOH 813-diOH-THC THC 11..OH.THC THC-COOH 8[[~IiOH-THC THC 11-OH-THC THC-COOH 8[~IiOH-THC I -* I Mean SEM N t THC 11-OH-THC THC-COOH 8~IiOH-THC THC 11-OI'I-THC THC-COOH 8I~IiOH-THC THC 11-OH-THC THC-COOH 8[~.diOH-THC Mean 21,5 72, SEM N THC 11-OH-THC THC-COOH 8[$.diOH-THC THC 11-OH-THC THC-COOH 81$diOH-THC THC 11-OH-THC THC-COOH 8[~-diOH-THC , ,0 74, , Mean SEM N ,8 480 THC 11-OH-THC THC-COOH 813-diOH-THC I Mean SEM 0.5 3, N * No specimen available. t N = number of specimens. 546

10 study. In addition, it appears that the smoking protocol employed in this work delivers a dose of THC that results in plasma THC concentration and ([THCCOOH] + [THC]) ratios similar to those found in other studies (21). Cannabinoids in urine McBumey et al. (19) reported that free 8~,11-diOH-THC was detected in urine, following enzymatic hydrolysis, within 4 h of smoking two marijuana cigarettes prepared to deliver a total dose of 150 ~g THC/kg body weight. These authors suggested that a concentration of this THC metabolite greater than 15 to 20 ng/ml might indicate marijuana use within 4--6 h of sample collection. Because THC and 11-OH-THC concentrations without prior glucuronide hydrolysis had been shown to be insignificant in urine after smoking (16), McBurney and others (16,19,36) 4.1 E v _J r- 35O o - loo - 5~ i 0 12o I _ 20 3O I I I I I I I T Urine THCCOOH Urine 25,~ I Urine THC 20 A 1.77% TI-IC I lo 5 2 ng/ml threshold, O, i i i i i T Time after smoking (rnin) Figure 3. Urine concentrations of THC and its major metabolites (mean + SEM) after smoking marijuana. suggested that 8~,ll-diOH-THC showed promise as a marker in urine for marijuana use within the time frame of previously reported pharmacologic effects. Only one subject in the present study (Tables VII and VIII) excreted 8]3,11-diOH-THC in quantities above the limit of detection of the assay for urine (0.9 ng/ml) (26). Following the 3.58% THC cigarette, 8~,11-diOH-THC was detected in the urine of subject #5 (Table VIII). The peak concentration was detected at 2 h post-administration, within the 0 to 4-h time frame reported by McBurney et al. (19). The addition of an enzymatic hydrolysis step in the extraction protocol, using bacterial J3-glucuronidase (from E. coli) demonstrated the presence of significant quantities of THC (up to 53.3 ng/ml) and 11-OH-THC (up to ng/ml) in urine (27). These finding do not support previously reported data that suggested that the concentrations of these neutral cannabinoids were "negligible" in human urine following marijuana smoking (16). Cannabinoid concentrations detected in the urine of 8 human subjects following marijuana smoking are individually presented in Tables VII and VIII for the 1.77% THC and 3.58% THC doses, respectively. Pharmacodynamic comparisons between doses are tabulated in Table IV. Mean urine profiles of THC, 11-OH-THC, and THC-COOH are presented in Figure 3. Unconjugated THC, liberated by the bacterial [~-glucuronidase, reached peak concentrations in urine ranging from 1.5 to 21.5 ng/ml (mean = 9.1 ng/ml) and from 3.2 to 53.3 ng/ml (mean = 21.5 ng/ml) for the 1.77% THC and 3.58% THC cigarettes, respectively (Figure 3). Peak THC levels in urine were reached 2 h after cessation of smoking in all subjects. Interindividual variation was high, a phenomenon reported in previous literature (19,37). This finding has been attributed in previous reports to intrasubject and intersubject differences in smoking efficiency (38), as well as to interindividual variations in the rate of metabolism of marijuana. It is very well documented that CYP450 (enzymes responsible for metabolizing cannabinoids and other zenobiotics) activities are influenced by many factors such as diet, age, and gender (39) and a number of other factors (e.g., fluid intake-output) (24). AUCs for the mean data (Table IV) were and ng-h/ml for the low and high dose, respectively, demonstrating a dose-dependent concentration relationship (p < 0.05). Urinary THC concentrations fell below the limit of detection of the assay (1.5 ng/ml at 5 h) for the 1.77% THC dose and 7 h for the 3.58% THC dose. In establishing mathematical models for the evaluation of recent marijuana use, Huestis et al. (24) used only plasma THC concentrations greater than or equal to 2 ng/ml. This proposed threshold allowed for the exclusion of very low residual concentrations of THC which may be present in plasma of frequent users even without the use of the drug in the very recent past (< 8 h). Using this same allowance of 2 ng/ml in the present study, our findings indicate that THC was only detectable in urine up to approximately 5 h after smoking the high dose (3.58% THC) of marijuana (Figure 3). This suggests that concentrations of THC in urine > 2 ng/ml, following enzymatic hydrolysis, would be indicative of marijuana use within 5 h of sample collection, well within the time frame in which behavioral and performance effects are known to occur (40,41). Bacterial enzymatic hydrolysis also allowed the detection and 547

11 quantitation of 11-OI-I-THC in urine, the primary psychoactive metabolite of THC (Figure 3). Mean peak concentrations were achieved at 3 h after the cessation of smoking (Figure 3). Peak concentrations (180 rain) ranged from 14.0 to 79.3 ng/ml with a mean ( SEM) of ng/ml for the 1.77% cigarette. For the 3.58% THC dose, the peak concentrations of ] 1-OH-THC ranged from 27.4 to 169 ng/ml with a mean ( SEM) of ng/ml. AUC for the mean 11-OH-THC data demonstrated the dose-dependent concentration relationship for concentrations (Table IV) of and ng-h/ml for the low and high dose, respectively (p < 0.05). 11-OH-THC was detectable 5 rnin after smoking, peaked at 3 h after smoking, then declined more gradually than THC, at a rate of 7.8 ng/ml/h and 12.7 ng/ml/h for the low and high doses, respectively. Concentrations of 11-OH-THC did not return to baseline by the end of the study (8 h). Urine concentrations of THC-COOH had the highest inter- and intrasubject variability and were detected for 8 h after smoking the 1.77% and 3.58% doses (Figure 3). Wide variability in the excretion of THC-COOH following marijuana smoking has been reported previously (12,19,42). Peak THC-COOH concentrations of ng/ml and ng/ml were detected 4 h after smoking for the low and high doses, respectively (Table IV). Mean THC-COOH concentrations measured at 4 h after smoking the placebo cigarette was ng/ml AUC for placebo, low, and high doses were , , and , respectively. Statistically significant differences for THC-COOH were not found between any of the three doses for mean peak concentration or AUC (Table IV). The wide interindividual variability of excretion and the persistence of moderate to high concentrations of THC-COOH over our 8-h study period indicate that this metabolite cannot be employed as a chemical marker in urine for recent use. Rather, THC-COOH in urine is only suggestive of marijuana use at some time in the past. This finding is in agreement with other studies which found that THC-COOH excretion was much too variable for characterizing patterns of marijuana use (19,33,34). Conclusions With the exception of one heavy cannabis user, 8[~-diOH-THC was not detected in the urine following marijuana smoking. In contrast to previously published research, this metabolite was not suggestive of recent marijuana use. Using bacterial enzymatic hydrolysis, however, THC and 11-OH-THC were identified and quantitated in urine. Urinary concentrations of THC greater than 2 ng/ml occurred within 5 h of smoking marijuana, a time during which psychological and performance effects are known to occur. These data suggest that THC may be a biologic marker for the identification of recent marijuana use. With the development and refinement of a procedure that facilitates the detection of free THC in urine by GC-MS, data must now be acquired on the excretion of this cannabinoid to more accurately determine what concentration represents residual levels from past use in chronic users. The current technology used for urine drug screening, initial immunoassay with a polyclonal antibody followed by GC-MS quantitation of THC-COOH, cannot be used to accurately predict the time of marijuana use or suggest any relationship between urine drug levels and human psychomotor performance. GC-MS testing of urine for THC, 11-OH-THC, and THC-COOH does provide additional information relating to the time of marijuana use and this information correlates with other published reports relating to psychomotor impairment after marijuana use can be used to determine performance impairment. References 1. N.R. Farnsworth. Pharmacognosy and chemistry of Cannabis safiva. J. Am. Pharm. Assoc. 9: (1969). 2. L. Lemberger. The metabolism of marijuana. Adv. Pharmacol. Chemother. 10: (1972). 3. National Household Survey On Drug Abuse: Main Findings 1998, Department of Health and Human Services, Substance abuse Administration, Office of Applied Statistics, 2000, Chapter 3. http;//drugabusestatistics.samsha.gov/, January SAMSHA, Office of Applied Studies, Mid-Year 200 Preliminary Emergency Department Data from the Drug Abuse Warning Network, National clearinghouse for Alcohol and Drug information (NCADI), Rockville, MD. http'j/drugabuse Statistics.samhsa.gov/, January lab emp_ drugtesting_index.html. 6. N. Benowitz. In Research findings on smoking abused substances, NIDA Research Monograph 89. C.N. Chiang and R.L. Hawks, Eds., 1990, pp M.A. Huestis, J.E. Henningfield, and E.J. Cone. Blood cannabinoids I. Absorption of THC and formation of 11 -OH-THC and THCCOOH during and after smoking marijuana. J. Anal. Toxicol. 16: (1992). 8. A. Ohlsson, J.E. Lindgren, A. Wahlen, S. Agurell, L.E. Hollister, and H.K. Gillespie. Plasma delta-9-tetrahydrocannabinol concentrations and clinical effects after oral and intravenous administration and smoking. Clin. PharmacoL Ther. 28: (1980). 9. D.M. Cocchetto, S.M. Owens, M. Perez-Reyes, S. Di Guiseppi, and L.L. Miller. Relationship between plasma delta-9-tetrahydrocannabinol concentration and pharmacologic effects in man. Psychopharmacology 75" (1981 ). 10. M.E. Wall, B.M. Sadler, D. Brine, H. Taylor, and M. Perez-Reyes. Metabolism, disposition, and kinetics of delta-9-tetrahydrocannabinol in men and women. Clin. Pharmacol. Ther. 34: (1983). 11. S. Agurell, M.M. Halldin, J. Lundgren, A. Ohlsson, M. Widman, H. Gillespie, and L. Hollister. Pharmacokinetics and metabolism of delta-l-tetrahydrocannabinol with emphasis on man. PharmacoL Rev. 38:21-43 (1986). 12. S. Agurell, M.M. Halldin, and L.E. Hollister: Pharmacokinetics and metabolism of Ag-tetrahydrocannabinol in man. In Biochemistry and Physiology of Substance Abuse, Vol. II, R. Watson, Ed. CRC Press, Boca Raton, FL, 1990, pp D.J. Harvey. Absorption distribution, and biotransformation of the cannabinoids. In Marihuana and Medicine, G.G. Nahas, K.M. Sutin, D.J. Harvey, and S. Agurell, Eds. Hurnana Press, Totowa, NJ, 2000, pp L.M. Bornheim and M.A. Correia. Purification and characterization of the major hepatic cannabinoid hydroxylase in the mouse: a possible member of the cytochrome P45011C subfamily. Mol. PharmacoL 40: (1991). 15. L.M. Bornheim, J.M. Lasker, and J.L. Raucy. Human hepatic microsomal metabolism of Al-tetrahydrocannabinol. Drug Metab. Dispos. 20: (1992). 548

12 16. M.E. Wall and M. Perez-Reyes. The metabolism of delta-9-tetrahydrocannabinol in man. J. Clin. Pharmacol. 21:178S-I 89S (I 981). 17. M. Widman, M.M. Halldin, and S. Agurell. Metabolism of delta- I -tetrahydrocannabinol in man. In Pharmacokinetics and Pharmacodynamics of Psychoactive Drugs, G. Barnett and C.N. Chiang, Eds. Biomedical Publications, Foster City, CA, 1985, pp W.W. Hanson, M.H. Buonarati, R.C. Basalt, N.A. Wade, C. Yep, A.A. Biasotti, V.C. Reeve, A.S. Wong, and N.W. Orbanosky. Comparison of 3H- and radioimmunoassay and gas chromatograph/mass spectrometry for the determination of delta-9-thc and cannabinoids in blood and serum. J. Anal. Toxicol. 7: (I 983). 19. L.J. McBurney, B.A. Bobble, and L.A. Sepp. GC/MS and EMIT analysis for A9-tetrahydrocannabinol metabolites in plasma and urine of human subjects. J. Anal. Toxicol. 10:56-63 (1986). 20. J.E. Manno, K.E. Ferslew, L.S. Franklin, and B.R. Manno. Human pursuit tracking performance (PTP) after smoking marihuana: correlation with plasma concentrations of delta-9-tetrahydrocannabinol (THC) and 11-nor-delta-9-tetrahydrocannabinol-9-carboxylic acid (nor-cooh THC). In Marihuana '84" Proceedings of the Oxford Symposium on Cannabis, D.J. Harvey, Ed. IRL Press, Oxford, U.K., 1985, pp M.A. Huestis, J.E. Henningfield and E.J. Cone. Blood cannabinoids I1. Models for the prediction of time of marijuana exposure from plasma concentrations of Ag-tetrahydrocannabinol (THC) and 11- nor-9-carboxy-ag-tetrahydrocannabinol (THCCOOH). J. Anal. ToxicoL 16: (1992). 22. A.D. Fraser and D. Worth. Urinary excretion profiles of 11-nor-9- carboxy-tetrahydrocannabinol: a A9-THCCOOH to creatine ratio. J. Anal, Toxicol. 23: (1999). 23. J.E. Manno, K.E. Ferslew, and B.R. Manno. Urine excretion patterns of cannabinoids and their clinical applications of the EMIT d.a.u. Cannabinoid urine assay for substance abuse treatment. In The Cannabinoids: Chemical, Pharmacologic and Therapeutic Aspects, S. Agurell, W.L. Dewey, and R.E. Willette, Eds. Academic Press, Orlando, FL, 1984, pp J.E. Manno. Interpretation of urinalysis results. In Urine Testing for Drugs of Abuse, NIDA Research Monograph 73, DHHS Publication Number (ADH) , R.L. Hawks and C.N. Chang, Eds. 1986, pp M.A. Huestis and E.J. Cone. Differentiating new marijuana use from residual drug excretion in occasional marijuana users. J. Anal. ToxicoL 22: (1998). 26. RM. Kemp, I.M. Abukhalaf, B.R. Manno, J.E. Manno, D.D. Alford, and G. Abusada. Cannabinoids in humans. L Analysis of Ag-tetrahy - drocannabinol and six metabolites in plasma and urine using GC-MS. J. Anal ToxicoL 19: (1995). 27. P.M. Kemp, I.K. Abukhalaf, B.R. Manno, i.e. Manno, and D.D. Alford. Cannabinoids in humans. II. The influence of three methods of hydrolysis on the concentration of THC and two metabolites in urine. J. Anal. Toxicol. 19: (1995). 28. M.A. EISohly and S. Feng. Ag-THC metabolite in meconium: Identification of 11 -OH-Ag-THC, 8~, 11 -dioh-ag-thc, and 11-nor- Ag-THC-9-COOH as major metabolites of Ag-THC. J. Anal. Toxicol. 22: (I 998). 29. R.L. Spitzer and J. Endicott. Schedule for Affective Disorders and Schizophrenia-Life-~me Version, 3rd ed., May '78-September '79. Biometrics Research Department, New York State Psychiatric Institute, New York. 30. R.J. Tallarida and R.B. Murray. Manual of Pharmacologic Calculations with Computer Programs, 2nd ed. Springer-Verlag, New York, NY, 1987, pp C.A. Hunt and R.T. Jones. Tolerance and disposition of tetrahydrocannabinol in man. J. PharmacoL Exp. Ther. 215:35-44 (1980). 32. E.W. Gill and G. Jones. Brain levels of delta-l-tetrahydrocannabinol and its metabolites in mice--correlation with behavior, and the effect of the metabolic inhibitors SKF 525A and piperonyl butoxide. Biochem. PharmacoL 21: (1972). 33. E. Johansson, H.K. Gillespie, and M.M. Halldin. Human urinary excretion profile after smoking and oral administration of [14C] A 9- tetrahydrocannabinol. J. Anal. Toxicol. 14" 176-I 80 (I 990). 34. C.A. Dackis, A.L.C. Pottash, W. Annitto, and M.S. Gold. Persistence of urinary marijuana levels after supervised abstinence. Am. J. Psychiatry 139:1196-II 98 (I 982). 35. P. Kelly and R.T. Jones. Metabolism of tetrahydrocannabinol in frequent and infrequent marijuana users. J. Anal. Toxicol. 16: (I 992). 36. G. Barnett, V. Licko, and T. Thompson Behavioral pharmacokinetics of marijuana. Psychopharmacology 85:51-56 (1985). 37. J.R. Soares and S.J. Gross. Separate radioimmunoassay measurements of body fluid delta-9-thc and 11-norcarboxy-delta-9-THC. LifeSci. 19: (1976). 38. A. Ohlsson, J.E. Lindgren, A. Wahlen, S. Agurell, L.E. Hollister, and H.K. Gillespie. Single dose kinetics of deuterium labeled Al-tetrahy - drocannabinol in heavy and light cannabis users. Biomed. Mass Spectrom. 9:6-10 (1982). 39. J.D. debethizy and J.R. Hayes. Metabolism: a determinant of toxicity. In Principles and Methods of Toxicology, 4th ed., A.W. Hayes, Ed. Taylor and Francis, Philadelphia, PA, 2001, pp L.E. Hollister. Health aspects of cannabis. Pharmacol. Rev. 38:1-20 (1986). 41. J.E. Manno, G.F. Kiplinger, S.E. Haine, I.F. Bennett, and R.B. Forney. Comparative effects of smoking marihuana or placebo on human motor and mental performance. Clin. Pharmacol. Ther. 11: (1970). 42. L.E. Hollister, H.K. Gillespie, A. Ohlsson, J.E. Lindgren, A. Wahlen, and S. Agurell. Do plasma concentrations of Ag-tetrahydro - cannabinol reflect the degree of intoxication? J. Clin. Pharmacol. 21: 171S-177S (1981 ). 549