Subjective, cognitive and cardiovasculardose-effect profile of nabilone and dronabinol inmarijuana smokers

Gillinder Bedi, Ziva D. Cooper & Margaret HaneyDivision on Substance Abuse, New York State Psychiatric Institute, and Department of Psychiatry, College of Physicians and Surgeons of Columbia University,New York, NY, USA

ABSTRACT adb_427 1..10

Marijuana dependence is a substantial public health problem, with existing treatments showing limited efficacy. Inlaboratory and clinical studies, the cannabinoid receptor 1 agonist oral D9tetrahydrocannabinol (THC; dronabinol) hasbeen shown to decrease marijuana withdrawal but not relapse. Dronabinol has poor bioavailability, potentially con-tributing to its failure to decrease relapse. The synthetic THC analogue, nabilone, has better bioavailability thandronabinol. We therefore aimed to characterize nabilone’s behavioral and physiological effects across a range of acutedoses in current marijuana smokers and compare these with dronabinol’s effects. Participants (4 female; 10 male)smoking marijuana 6.6 (standard deviation = 0.7) days/week completed this outpatient, within-subjects, double-blind, randomized protocol. Over seven sessions, the time-dependent subjective, cognitive and cardiovascular effects ofnabilone (2, 4, 6, 8 mg), dronabinol (10, 20 mg) and placebo were assessed. Nabilone (4, 6, 8 mg) and dronabinol (10,20 mg) increased ratings of feeling a good effect, a strong effect and/or ‘high’ relative to placebo; nabilone had a sloweronset of peak subjective effects than dronabinol. Nabilone (6, 8 mg) modestly lowered psychomotor speed relative toplacebo and dronabinol. There were dose-dependent increases in heart rate after nabilone, and nabilone (2 mg) anddronabinol (10 mg) decreased systolic blood pressure. Thus, nabilone produced sustained, dose-related increases inpositive mood, few cognitive decrements and lawful cardiovascular alterations. It had a longer time to peak effects thandronabinol, and effects were more dose-related, suggesting improved bioavailability. Nabilone was well tolerated bymarijuana smokers, supporting further testing as a potential medication for marijuana dependence.

Keywords Agonist treatment, dose-effect profile, dronabinol, marijuana dependence, nabilone.

Correspondence to: Margaret Haney, New York State Psychiatric Institute, College of Physicians and Surgeons of Columbia University, 1051 RiversideDrive, Unit 120, New York, NY 10032, USA. E-mail: mh235@columbia.edu

INTRODUCTION

Marijuana is the most commonly used illegal drug in theUnited States (SAMSHA 2010) and internationally(United Nations Office on Drugs and Crime 2007).Although most people who initiate marijuana use do notdevelop problematic use of the drug (Brook, Zhang &Brook 2011), a subset progress to smoking marijuanadaily (around 8% of past year users; Chen, Kandel &Davies 1997). A sizable minority (some 20%) of daily ornear-daily marijuana smokers meet criteria for mari-juana dependence (Chen et al. 1997). Moreover, uponcessation of daily marijuana use, a characteristic constel-lation of withdrawal symptoms, including craving,

anxiety, irritability and sleep and appetite disruptions,has been demonstrated to occur in both laboratory(Haney et al. 1999b, 2004, 2008) and clinical (Vandreyet al. 2008) settings. Thus, although the proportion ofmarijuana users who experience dependence is small, thehigh prevalence of marijuana smoking means that theabsolute number of individuals with problematic use ofthe drug is greater than that for illegal drugs with higherabuse liability (Budney et al. 2007a; SAMSHA 2010).

Consistent with this, estimates indicate that some1.2 million people in the United States received treatmentfor marijuana use in 2009 (SAMSHA 2010). Existingtreatments are primarily behavioral and psychotherapeu-tic, and have been reported to assist in reducing

ORIGINAL ARTICLE

Addiction Biologydoi:10.1111/j.1369-1600.2011.00427.x

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology

marijuana use and increasing cessation rates (MTPResearch Group 2004; Budney et al. 2006; Kadden et al.2007). However, as is the case for other drugs, relapserates remain high. For example, one study assessing theefficacy of voucher-based contingency management,alone and combined with cognitive behavioral therapy(CBT), reported that only 17% (vouchers alone) to 37%(vouchers plus CBT) of patients were abstinent 12months after treatment cessation (Budney et al. 2006).To better address the needs of treatment-seeking mari-juana smokers, it is clear that new treatments andadjuncts to existing treatments are needed. Combinationsof psychotherapy and pharmacotherapy are most effec-tive in treatment of primary psychiatric conditions suchas depression (Zobel et al. 2011) and such strategiesenhance treatment for other drugs such as cigarettes(Grassi et al. 2011) and heroin (Nunes et al. 2006), sug-gesting that a similar approach may improve treatment ofmarijuana dependence. However, there is, as yet, no Foodand Drug Administration (FDA)-approved medication toaid in cessation or reduction of marijuana use.

One approach successfully employed in treating opiateand tobacco dependence is the use of agonist medica-tions to alleviate withdrawal symptoms that may pre-cipitate relapse (Threlkeld et al. 2006; Stapleton &Sutherland 2011). The cannabinoid receptor type 1and 2 (CB1 and CB2) agonist dronabinol [syntheticD9tetrahydrocannabinol (THC); Marinol®, Abbott Phar-maceuticals, Inc, Abbott Park, IL, USA] reduces with-drawal symptoms in abstinent marijuana smokers (Haneyet al. 2004; Budney et al. 2007b; Levin et al. 2011),however dronabinol alone did not affect relapse rates inlaboratory (Haney et al. 2008) or clinical (Levin et al.2011) studies. Dronabinol has low bioavailability (from 4to 20%; McGilveray 2005; Ben Amar 2006), and its effectsdo not reliably show clear dose dependency (e.g. see Haneyet al. 1999a), perhaps contributing to its lack of effect onmarijuana relapse. A substantial further limitation is thaturinary metabolites of dronabinol are identical to those ofmarijuana (Lile, Kelly & Hays 2010), preventing differen-tiation between compliance with medication and ongoingmarijuana use in clinical and research settings.

An alternative to dronabinol is nabilone (Cesamet®,Meda Pharmaceuticals, Somerset, NJ, USA), a syntheticcannabinoid analogue of THC that is FDA-approved fornausea and emesis in chemotherapy patients. Nabiloneproduces urinary metabolites distinct from those of mari-juana (Fraser & Meatherall 1989). In addition, nabilone ismore potent than dronabinol, has greater bioavailability(at least 60%; Lemberger et al. 1982; Ben Amar 2006)and demonstrates dose linearity (Lemberger et al. 1982).Two recent drug discrimination studies have tested arange of dronabinol and nabilone doses in a marijuana-using sample. One study (n = 6) showed that nabilone

(2, 3 and 5 mg) and dronabinol (10, 15 and 25 mg) sub-stituted for the test dose of dronabinol (25 mg), indicatingsubstantial overlap in the interoceptive properties of thetwo drugs (Lile et al. 2010). Dronabinol and nabilone alsoproduced lawful increases in marijuana-like subjectiveeffects such as feeling a good drug effect, feeling ‘high’ andfeeling ‘stoned’, as well as increasing heart rate. Thesecond study (n = 6) demonstrated that nabilone (1, 3 mg)pre-treatment enhanced the discriminative-stimuluseffects of dronabinol (5, 10, 15 mg), demonstrating thesafety of combining these two orally administered can-nabinoid agonists (Lile, Kelly & Hays 2011).

The purpose of the present study was to compare thedose-effect profile and time course of the effects ofnabilone and dronabinol in volunteers who smoke mari-juana heavily (repeatedly each day, 6–7 days/week), asthis is the most relevant population in which to studypotential treatment approaches for marijuana depen-dence (Haney 2008). Because regular marijuanasmokers are tolerant to cannabinoid effects (Hart et al.2001; Haney 2007), we administered higher nabilonedoses than have been previously tested. Characterizingthe effects of a range of doses of nabilone relative to dron-abinol on (1) self-reported mood state; (2) cognitivefunction; and (3) cardiovascular function will provideessential information about appropriate doses to employto assess the effects of nabilone on marijuana withdrawaland relapse.

MATERIALS AND METHODS

Participants

Healthy male and female regular marijuana smokerswere recruited using newspaper and Internet advertise-ments. Participants were required to (1) be between 21and 45 years of age; (2) state that they were not seekingtreatment for marijuana use; and (3) agree to use effec-tive birth control for the duration of the study (females).Exclusion was based on (1) Axis 1 psychiatric disorderrequiring medical intervention (Diagnostic and Statisti-cal Manual of Mental Disorders 4th Edition criteria; APA1994); (2) inability to abstain from drugs other thanmarijuana for the study duration, verified by urine toxi-cology; (3) medical findings counterindicating participa-tion; (4) inability to perform study procedures; and (5)current pregnancy or lactation (female candidates). Priorto enrollment, candidates underwent comprehensivemedical and psychiatric screening including electrocar-diogram, urine and blood toxicology, and clinicalinterview. At enrollment, participants provided writteninformed consent. At study completion, they underwentfull debriefing and received compensation for participa-tion. All study procedures were approved by the New York

2 Gillinder Bedi et al.

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology

State Psychiatric Institute’s (NYSPI) Institutional ReviewBoard and were in accordance with the Helsinki Declara-tion of 1964, revised in 2008.

Experimental protocol

This outpatient study employed a within-subjects, ran-domized, double-blind design. Participants initially under-went a practice session in which no medication wasadministered to become familiar with study procedures.They then completed seven 6-hour drug-administrationsessions separated by at least 48 hours to allow for studymedication clearance. Sessions took place in a comfortablelaboratory environment; participants were tested indi-vidually to prevent effects of social interaction with otherparticipants on responses to the study medications. Par-ticipants were instructed to not eat breakfast on themorning of sessions and to not consume alcohol for 24hours prior to session initiation. A breathalyzer test uponarrival confirmed compliance with this requirement; par-ticipants also provided samples for urine toxicologyscreens to ensure no recent use of drugs other than mari-juana. Participants were required to abstain from smokingmarijuana or cigarettes from midnight the night prior tosessions. To confirm compliance, breath carbon monoxidewas tested and was required to be �8 p.p.m. prior to com-mencement of the session. Female participants’ urine wasalso screened for pregnancy twice during participation,which was approximately 3 weeks in duration. After thesetests, participants consumed a light breakfast of theirchoice that was standardized across sessions. Baselinesubjective, cognitive and cardiovascular measurements(see below) were collected. Study medications wereadministered around 60 minutes after arrival. Subjective,cognitive and cardiovascular measurements were col-lected at 30-minute intervals for 2 hours following medi-cation administration; thereafter measurements weretaken hourly for 3 hours. Cigarette-smoking participantswere allowed to smoke cigarettes at the same time in eachsession to prevent the onset of nicotine withdrawal overthe 6-hour session (Parrott et al. 1996). Following finalassessments, participants were required to pass a stan-dardized field sobriety test prior to leaving the laboratory.They were instructed not to drive a car for 8 hours follow-ing sessions and were reimbursed for the cost of publictransportation to and from each session.

Over the seven sessions, participants received nabilone(2, 4, 6, 8 mg), dronabinol (10, 20 mg) and placebo inrandomized order. Nabilone (0, 1 mg; Meda Pharmaceu-ticals Inc, Somerset, NJ, USA) and dronabinol (10 mg;Unimed Pharmaceuticals, Buffalo Grove, IL, USA) cap-sules were overencapsulated by the NYSPI Pharmacy insize 00 capsules with lactose filler. Because one condition(nabilone 8 mg) required that a total of 8 capsules (1 mg

nabilone each) be administered, 8 capsules were admin-istered in each session with the number of placebo cap-sules varying according to active condition. The lowestnabilone dose selected is equivalent to the upper limit ofthe recommended dose for nausea; doses higher than this(4, 6 and 8 mg) were also employed because the regularmarijuana smokers recruited were expected to show tol-erance to cannabinoid effects (Hart et al. 2001; Haney2007). We have previously shown doses of 10 and 20 mgof dronabinol to be well tolerated in marijuana smokers(Hart et al. 2005), and repeated doses of 10 and 20 mgdronabinol alleviate withdrawal symptoms in abstinentmarijuana smokers (Haney et al. 2004; Budney et al.2007b; Levin et al. 2011).

Outcome measures

Subjective mood and drug effect

Subjective effects of the capsules were measured using aseries of Visual Analogue Scales (VAS; Folstein & Luria1973) and a Capsule Ratings Form (CRF; Haney 2007),which we have previously found to be sensitive to theeffects of oral cannabinoids (Haney 2007; Haney et al.2008; Bedi et al. 2010). The VAS consisted of 44 descrip-tors assessing mood (e.g. ‘content’), medication effects(e.g. ‘bad effect’) and physical symptoms (e.g. ‘blurredvision’). At each assessment, participants were requiredto rate the extent to which each descriptor applied tothem at that time on a 100 mm computer-presentedvisual analogue scale anchored with ‘not at all’ and‘extremely’. To limit the number of analyses conducted,we reduced 35 of the 44 items into 7 cluster scales, cal-culated as arithmetic means of individual items based ona previously conducted factor analysis (see Haney 2007).The seven clusters employed were (1) High (consisting ofthe descriptors ‘good drug effect’ and ‘high’); (2) Miser-able (’depressed’, ‘irritable’ and ‘miserable’); (3) Tired(‘clumsy’, ‘sedated’, ‘sleepy’, ‘withdrawn’, ‘unmotivated’and ‘tired’); (4) On Edge (‘anxious’, ‘jittery’, ‘restless’,‘stimulated’ and ‘on edge’); (5) Bad Effect (‘blurredvision’, ‘chills’, ‘dizzy’, ‘headache’, ‘heart pound’, ‘nau-seous’, ‘stomach pain’, ‘upset stomach’ and ‘bad drugeffect’); (6) Content (‘alert’, ‘energetic’, ‘mellow’, ‘self-confident’, ‘social’, ‘talkative’ and ‘content’); and (7) Con-fused (‘can’t concentrate’, ‘forgetful’ and ‘confused’). Inaddition to these cluster scales, we also included two indi-vidual items that were not included in the clusters:‘hunger’ and ‘marijuana craving’.

The CRF, presented in paper and pencil format, con-sisted of five items assessing the extent to which partici-pants: felt a good drug effect, felt a bad drug effect, felt astrong effect, liked the effect they felt and would be willingto take the drug again. Each item was rated on a 100 mmVAS.

Nabilone in marijuana smokers 3

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology

Cognition

Cognitive function was assessed with a computerized per-formance battery including: a Digit Span immediate anddelayed recall and recognition task (DIG; 2 minutes); aDivided Attention Task (DAT; 10 minutes); a Digit SymbolSubstitution Task (DSST; 3 minutes); and a RepeatedAcquisition Task (RAT; 3 minutes). This battery, designedfor repeated assessments, measures basic as well ashigher-level cognitive functions including psychomotorspeed, simple and divided attention, and verbal andworking memory. We have previously shown this batteryto be sensitive to the effects of oral cannabinoids (Hartet al. 2005; Bedi et al. 2010).

Cardiovascular function

Heart rate and seated blood pressure were measured witha Sentry II vital signs monitor (Model 6100, NBS MedicalServices, Costa Mesa, CA, USA). In addition, becausenabilone can produce orthostatic hypotension (Lem-berger et al. 1982), we repeatedly assessed change in sys-tolic and diastolic blood pressure from sitting to standingduring each session; no participant was allowed to leave asession until vital signs, including orthostatic blood pres-sure, were normal.

Data analyses

The outcome variables for subjective (VAS and CRFscores), cognitive (DSST, DAT, DIG and RAT scores) andphysiological (heart rate, blood pressure) measures werecalculated as the peak score of the six (cognitive tests) orseven (subjective and cardiovascular assessments) post-baseline measurements taken over each session. Prior tomain analyses, we assessed for differences in baselinescores across sessions for all variables except the CRF,which was not employed at baseline because no capsulehad been administered; no significant differences wereobserved. Effects of drug condition were assessed withsingle-factor repeated measures Analyses of Variance(ANOVAs) followed by planned contrasts comparing: (1)nabilone 2, 4, 6 and 8 mg with placebo; (2) dronabinol10 and 20 mg with placebo; (3) nabilone 2, 4, 6 and8 mg with dronabinol 10 mg; and (4) nabilone 2, 4, 6and 8 mg with dronabinol 20 mg. With this approach,we aimed to assess both the effects of the active drugconditions relative to placebo and to compare thenabilone and dronabinol conditions. Alpha was set atP < 0.05. Where no significant departure from sphericitywas indicated by Mauchly’s test of sphericity (P > 0.05),planned contrasts were calculated using the omnibusANOVA residual mean squares and degrees of freedom(see Sheskin 2000). In the case that a significant depar-ture from sphericity was observed, we conservatively

employed the individual contrast error term (Keppel1991). We observed isolated univariate outliers(z > � 3.29) across measures. In all except two cases,removal of the outlier did not affect outcomes and origi-nal data were retained. In two outcome measures (VASConfused cluster and CRF Bad Drug Effect) removal of theoutlying data point altered the results; therefore resultsare presented with data from one participant removed.Analyses were undertaken using SPSS 18.0 (SPSS,Chicago, IL, USA).

RESULTS

Fourteen participants (4 female; 10 male) completed thestudy. A further five volunteers (3 female; 2 male) startedbut did not complete the protocol. Of these, one droppedfrom the study reporting that she disliked the medicationeffects, one reportedly felt uncomfortable in the labora-tory environment, one was dropped after twice producingurine samples positive for benzodiazepines, and two didnot attend sessions as scheduled.

Of participants who completed the study, four identi-fied as white (one was Hispanic), nine were black (threewere Hispanic) and one was reportedly of mixed racialorigin. They had a mean age of 27.9 [standard deviation(SD) = 5.1] years and reported an average of 13.1(SD = 2.5) years of education. Participants reportedlysmoked marijuana 6.6 (SD = 0.7) days per week,smoking 6.9 (SD = 4.5) ‘joints’ (marijuana cigarettes) perday at a weekly cost of $80.50 (SD = 59.8). On average,they reported having first smoked marijuana at 13.9(SD = 3.5) years of age and had smoked the drug regu-larly for 9.9 (SD = 6.7) years. Twelve reported somecurrent alcohol consumption, drinking on average 2.8(SD = 1.3) alcoholic drinks on 1.2 (SD = 1.2) days perweek. Nine were current cigarette smokers, smoking 4.4(SD = 3.2) cigarettes per day.

Subjective drug effects and mood state

The higher doses of nabilone were rated as producingpositive drug effects relative to placebo. As shown inFig. 1, nabilone (6, 8 mg) increased CRF ‘good drugeffect’ relative to placebo, whereas the lowest dose ofnabilone (2 mg) produced lower peak ratings on thismeasure compared with dronabinol (10 mg). Figure 2shows a similar pattern for the VAS ‘high’ cluster; thehighest nabilone doses (6, 8 mg) increased ratings com-pared with placebo; nabilone (6 mg) also producedhigher ratings relative to dronabinol (20 mg). On boththe CRF ‘good drug effect’ and VAS ‘high’ clusters,the intermediate dose of nabilone (6 mg) produced thehighest ratings at 39.7/100 � 10.8 and 39.5/100 � 10.5, respectively, revealing an inverted u-shapedfunction.

4 Gillinder Bedi et al.

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology

On the CRF ‘strong drug effect’ measure (Fig. 1), thethree higher doses of nabilone (4, 6, 8 mg) as well as bothdoses of dronabinol increased scores compared withplacebo. Nabilone (2 mg) produced lower ratings on‘strong drug effect’ than dronabinol (10 mg). The highestdoses of nabilone (6, 8 mg) significantly increased CRFratings of drug likeability and willingness to take the drugagain relative to placebo. Nabilone (8 mg) also increased

ratings of willingness to take the drug again comparedwith dronabinol (20 mg), with no other differencesobserved between nabilone and dronabinol on these mea-sures (see Table 1).

Nabilone (6 mg) increased ratings on the VAS ‘tired’cluster relative to placebo, whereas both higher doses ofnabilone (6, 8 mg) increased ‘tired’ ratings comparedwith the lower dose of dronabinol (10 mg). Nabilone (6,8 mg) increased ratings on the VAS ‘on edge’ cluster com-pared with dronabinol (10 mg); nabilone (6 mg) alsoincreased scores on this measure relative to dronabinol(20 mg). However, ratings of ‘tired’ and ‘on edge’ werelow with no drug condition producing peak scores higherthan 17 out of a possible 100 on either of these measures(see Table 1).

There were no adverse reactions requiring medicalintervention observed during the study and no effect ofdrug condition on the majority of negative affect andphysical symptom measures (other than small changes in‘tired’ and ‘on edge’). Neither nabilone nor dronabinolaffected ratings of ‘bad drug effects’, or feeling ‘miser-able’, ‘confused’, ‘content’ or ‘hungry’ (data not pre-sented). There was also no effect of medication on VASratings of craving for marijuana (see Table 1).

As shown in Fig. 2, nabilone had a longer time to peaksubjective effects than did dronabinol. Peak effects on VAS

Figure 1 Effects of nabilone (2, 4, 6, 8 mg) and dronabinol (10,20 mg) on self-rated good drug effect (top) and strong drug effect(bottom) in regular marijuana smokers. Data are means of peakscores from seven assessments taken after capsule administration(� SEM; maximum possible score = 100). Asterisks denote a signifi-cant difference from placebo (P < 0.05) in the indicated condition. $denotes a significant difference from dronabinol (10 mg). CRF =capsule rating form

Figure 2 Time course of effects of nabilone (2, 4, 6, 8 mg; top) anddronabinol (10, 20 mg; bottom) on self-rated feelings of being high inregular marijuana smokers. Data are means (�SEM; maximum pos-sible score = 100).VAS = visual analogue scale

Nabilone in marijuana smokers 5

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology

‘high’ ratings were recorded at 120 minutes for nabilone(2 mg), 180 minutes for both the 4 and 8 mg doses, and240 minutes post-administration for nabilone (6 mg).Conversely, peak effects for both doses of dronabinoloccurred 90 minutes after capsule administration, con-sistent with earlier findings (e.g. Haney et al. 2005). Asimilar pattern, of nabilone producing later peak subjec-tive effects than dronabinol, was observed on other mea-sures such as CRF ‘strong drug effect’.

Cognitive function

There were limited effects of nabilone on the range ofcognitive functions assessed. As shown in Fig. 3, nabilonereduced psychomotor speed relative to placebo and dron-abinol, with nabilone (6, 8 mg) decreasing the totalnumber of items attempted on the DSST compared with

placebo and dronabinol (10 mg), and nabilone (6 mg)decreasing the number attempted compared with dron-abinol (20 mg). Nabilone (6, 8 mg) also decreased thenumber of items correct on the DSST compared withplacebo; nabilone (6 mg) decreased DSST scores com-pared with both doses of dronabinol (Table 1). The higherdoses of nabilone (6, 8 mg) decreased the percentagecorrect on the DSST relative to placebo and dronabinol(20 mg); nabilone 4 mg decreased DSST percent correctcompared with dronabinol only (20 mg; see Table 1).

The number of false alarms on the DAT was higherafter dronabinol administration (10 mg) than nabilone(8 mg; Table 1) and the total tracking distance on theDAT was lower after nabilone (2 mg) and dronabinol(10 mg) than placebo (see Table 1). There was no effect ofmedication condition on the number of hits, hit latencyor maximum tracking speed on the DAT, the number of

Table 1 Effects of nabilone and dronabinol on peak scores for subjective mood state and cognitive function.

PBO NAB2 NAB4 NAB6 NAB8 DRON10 DRON20

Significant plannedcomparisons*

Meana Meana Meana Meana Meana Meana Meana

(SEM) (SEM) (SEM) (SEM) (SEM) (SEM) (SEM)

CRF Capsule likingb 21.7 17.0 32.8 38.5 38.5 29.8 28.4 NAB6 > PBO(9.7) (8.2) (9.7) (10.0) (8.6) (8.8) (10.4) NAB8 > PBO

CRF Would takecapsule againb

24.1 15.2 30.3 40.5 34.1 29.3 26.5 NAB6 > PBO(9.4) (8.1) (10.3) (11.3) (9.9) (9.8) (9.9) NAB8 > PBO

NAB8 > DRON20VAS Tiredb 10.1 8.8 11.5 16.1 13.0 6.8 12.0 NAB6 > PBO

(2.9) (1.9) (2.8) (4.1) (2.5) (1.7) (3.9) NAB6 > DRON10NAB8 > DRON10

VAS On edgeb 9.0 8.4 9.2 12.2 10.4 6.3 7.3 NAB6 > DRON10(2.9) (3.2) (2.2) (3.5) (3.3) (2.4) (3.2) NAB8 > DRON10

NAB6 > DRON20VAS Marijuana

cravingb,c

63.6 58.6 64.8 63.9 64.9 64.9 55.5 –(8.5) (7.7) (9.9) (9.3) (9.4) (9.4) (11.2)

DAT Number offalse alarms

9.3 5.9 6.5 6.2 4.5 7.3 4.0 DRON10 > NAB8(4.1) (2.3) (2.3) (1.5) (1.3) (2.0) (1.4)

DAT Total Trackingdistance

19692.4 15414.1 14195.1 16172.5 18696.2 16064.9 18549.5 PBO > NAB2(4721.1) (3480.3) (2082.1) (3569.4) 3125.1 (3618.3) (4489.3) PBO > DRON10

DSST Total Correct 91.2 88.4 87.0 81.6 85.4 89.6 88.4 PBO > NAB6(3.8) (3.3) (2.8) (3.6) (2.6) (3.6) (3.3) PBO > NAB8

DRON10 > NAB6DRON20 > NAB6

DSST % Correct 98.7 98.4 98.2 97.8 98.1 98.2 98.9 PBO > NAB6(0.5) (0.4) (0.8) (0.6) (0.9) (0.6) (0.4) PBO > NAB8

DRON20 > NAB4DRON20 > NAB6DRON20 > NAB8

RAT Number oferrors

57.2 55.1 50.9 48.9 53.0 54.7 50.3 PBO > NAB4(5.5) (4.2) (4.8) (4.7) (5.2) (5.9) (4.7) PBO > NAB6

PBO > DRON20

*P < 0.05. aIndividual subject peak scores across the seven (or six for cognitive measures) time points after capsule administration were calculated; dataare means of these peak scores. bScores range from 0 to 100. cn = 13 due to missing data. CRF = capsule rating form; DAT = divided attention task;DRON10 = dronabinol 10 mg; DRON20 = dronabinol 20 mg; DSST = digit symbol substitution task; NAB2 = nabilone 2 mg; NAB4 = nabilone 4 mg;NAB6 = nabilone 6 mg; NAB8 = nabilone 8 mg; PBO = placebo; RAT = rapid acquisition task; SEM = standard error of the mean; VAS = visual analoguescale.

6 Gillinder Bedi et al.

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology

items entered in the Digit Span task, or on immediate ordelayed recall or recognition scores (data not presented).

On the RAT, which taps, among other things, workingmemory, there was no effect of medication on the totalnumber of trials completed (data not presented). Indeed,the number of errors made in this task was higher onplacebo relative to nabilone (4, 6 mg) and dronabinol(20 mg; Table 1).

Cardiovascular function

As shown in Fig. 4, nabilone produced dose-dependentincreases in heart rate: heart rate following nabiloneadministration (4, 6, 8 mg) was significantly higher thanwhen placebo or dronabinol (10, 20 mg) were given.Nabilone (2 mg) and dronabinol (10 mg) resulted indecreases in seated systolic blood pressure relative toplacebo. There was no effect of medication condition onseated diastolic blood pressure (data not presented). Of686 measurements of standing versus seated blood pres-sure taken after drug administration, 11 met criteria fororthostatic hypotension (>20 mmHg decrease in systolicblood pressure and/or >10 mmHg decrease in diastolicblood pressure upon standing). Of these measurements,two occurred after placebo administration, one afternabilone (4 mg), two after nabilone (6 mg), four afternabilone (8 mg) and two occurred after administration of

dronabinol (20 mg). None of these instances requiredmedical intervention.

DISCUSSION

Nabilone, at acute doses higher than administered in pre-vious studies (6 and 8 mg compared with 5 mg, thehighest dose previously administered to marijuanasmokers; Lile et al. 2010) and up to four times higherthan doses given clinically for nausea, was well toleratedin this marijuana-using sample, as indicated by

Figure 3 Effects of nabilone (2, 4, 6, 8 mg) and dronabinol (10,20 mg) on psychomotor speed in regular marijuana smokers. Dataare means of peak scores from six assessments taken after capsuleadministration (� SEM). Asterisks denote a significant differencefrom placebo (P < 0.05) in the indicated condition. $ denotes asignificant difference from dronabinol (10 mg). + denotes a significantdifference from dronabinol (20 mg)

Figure 4 Effects of nabilone (2, 4, 6, 8 mg) and dronabinol (10,20 mg) on heart rate (top) and systolic blood pressure (bottom) inregular marijuana smokers. Data are means of peak scores fromseven assessments taken after capsule administration (�SEM).Aster-isks denote a significant difference from placebo (P < 0.05) in theindicated condition. $ denotes a significant difference from dronab-inol (10 mg). + denotes a significant difference from dronabinol(20 mg). BP = blood pressure

Nabilone in marijuana smokers 7

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology

dose-related increases in self-reported positive drugeffects and few negative effects. Nabilone caused orderlyincreases in heart rate that did not require medical inter-vention in this healthy sample. Although the increasesobserved on the higher dose of nabilone (8 mg), which insome individuals exceeded 100 beats per minute (b.p.m.;up to 119 b.p.m.) could be medically significant in a clini-cal population, on nabilone (4 mg) no individual’s heartrate exceeded 100 b.p.m. at any time point, and nabilone(6 mg) produced only one reading over 100 b.p.m.(106 b.p.m.), suggesting limited cardiovascular risk atthese doses. Effects on cognitive function were minor,with the main finding being modest decreases in psycho-motor speed at the highest doses of nabilone (6, 8 mg).

The lowest nabilone dose (2 mg) had few effectsoverall, whereas the 4 mg dose most closely mirroreddronabinol’s (10, 20 mg) effects; these findings are con-sistent with earlier findings showing that nabilone (3,5 mg) substituted for dronabinol (25 mg) in a drug dis-crimination paradigm (Lile et al. 2010). In the currentstudy, higher doses of nabilone (6, 8 mg) produced themost robust positive subjective effects, disruption in cog-nitive performance and cardiovascular changes. Impor-tantly, the positive drug effects produced by both nabiloneand dronabinol were significant, but even these highdoses did not compare with the peak ratings of ‘high’produced by three puffs of low-potency smoked mari-juana (e.g. Haney et al. 1999b; Hart et al. 2001; Haney2002). In addition, the effects of nabilone were markedlymore sustained than the effects of smoked marijuana(Haney et al. 1999b; Hart et al. 2001; Haney 2002), sup-porting its possible utility as a treatment for marijuanadependence. This is consistent with earlier reports of pro-longed effects of nabilone in healthy volunteers (Lem-berger & Rowe 1975). In the present study, the effects ofnabilone were still apparent 5 hours after capsule admin-istration. Although the session length was insufficient todetermine whether nabilone’s effects were longer lastingthan those of dronabinol, observation of the data (e.g.see Fig. 2) suggest that this might be the case. Futureresearch should address this possibility.

The present results are broadly consistent with earlierstudies investigating the effects of nabilone in humans,after accounting for methodological differences. A priorlaboratory study indicated that a low dose of nabilone(2 mg), while producing modest increases in subjectivereports of feeling ‘high’, did not increase ‘elation’ ratingsin daily, weekly or occasional marijuana smokers (Men-delson & Mello 1984), consistent with the present find-ings that the lower dose of nabilone did not producepositive mood effects. Whereas an earlier study reportedthat lower doses of nabilone (1–3 mg) produced markedcognitive impairments (Wesnes et al. 2010), we foundonly modest decreases in psychomotor and potentially

information processing speed. The most parsimoniousexplanation for this difference is that our sample con-sisted of near-daily marijuana smokers who would beexpected to demonstrate tolerance to the effects of can-nabinoids, whereas the earlier study employed non-marijuana smoking volunteers. We have previouslyreported the cognitive effects of smoked and oral cannab-inoids in regular marijuana smokers to be mild (Hartet al. 2001, 2010; Bedi et al. 2010). This earlier study(Wesnes et al. 2010) also reported common and some-times severe adverse effects of nabilone such as highlevels of anxiety and aggression, and loss of conscious-ness. In the present study, however, the doses employeddid not produce substantial adverse effects, despite beingalmost three times higher than the highest dosesemployed by Wesnes et al. (2010), further supporting tol-erance to these effects in our clinically relevant sample.

Although the present study did not measure the rein-forcing effects of nabilone or dronabinol, a recent reviewof evidence from Canada, where nabilone has beenapproved for some 30 years for chemotherapy-relatednausea, failed to reveal substantial evidence of diversionand concluded that nabilone abuse appears to be rare(Ware and St Arnaud-Trempe 2010). Moreover, in ahuman laboratory study evaluating the reinforcingeffects of a low dose of nabilone (2 mg), no operantresponding was observed for nabilone by daily, weekly oroccasional marijuana smokers (Mendelson & Mello1984). However, it is noteworthy that this prior study(Mendelson & Mello 1984) examined a dose of nabilone(2 mg) that in the present study did not produce positivemood effects. Thus, it is possible that higher doses ofnabilone may be more readily self-administered. Our find-ings that the 6 and 8 mg doses of nabilone increasedratings of drug likeability and willingness to take the drugagain would support direct investigation of their reinforc-ing effects in regular marijuana users.

There are numerous other avenues for futureresearch. It remains to be seen whether nabilonedecreases withdrawal symptoms in abstinent heavy mari-juana users, as has been demonstrated to be the case fordronabinol (Haney et al. 2004; Budney et al. 2007b;Levin et al. 2011). In the present study, there was noeffect of nabilone on craving for marijuana (see Table 1)however participants had been abstinent for fewer than24 hours. Marijuana withdrawal symptoms typicallyemerge after some 24 hours of abstinence, suggestingthat nabilone’s effects on withdrawal symptoms, includ-ing craving, might require a longer period of abstinenceto become apparent. We have previously reported thatdronabinol decreases marijuana craving during with-drawal (Haney et al. 2004), suggesting that nabilone mayhave a similar effect. A prior study (Lile et al. 2011) hasshown that lower doses of nabilone (1, 3 mg) are well

8 Gillinder Bedi et al.

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology

tolerated in combination with dronabinol (5, 15, 30 mg).Future studies should determine how nabilone influencesthe direct effects of smoked marijuana as well as hownabilone influences marijuana withdrawal and relapse.These data can then inform clinical testing of thisapproach in treatment-seeking marijuana users.

Although dronabinol has shown promise in reducingwithdrawal symptoms in abstinent marijuana users(Haney et al. 2004; Budney et al. 2007b; Levin et al.2011), to date it has not proven effective in reducingrelapse rates (Haney et al. 2008; Levin et al. 2011). Theabsence of clear dose-dependence in dronabinol’s effectsin this and earlier studies (e.g. Haney et al. 1999a), whichmay relate to its variable metabolism and low bioavail-ability (McGilveray 2005; Ben Amar 2006), may havecontributed to its failure to reduce relapse rates in studiesthus far. Nabilone, which is better absorbed, has a longertime to peak effects and has urinary metabolites that aredistinct from those of marijuana, may thus be a superiormedication with which to investigate the clinical utility ofagonist treatment in marijuana dependence. The presentdata support nabilone as a promising candidate medica-tion and demonstrate the importance of further investi-gation of nabilone as a potential treatment for marijuanadependence.

Acknowledgements

The authors would like to thank Michael Harakas forassistance in data collection. This research was supportedby the National Institute on Drug Abuse (DA09236). Thisstudy complies with the laws of the country in which itwas performed (United States).

Disclosure/Conflict of Interest

The authors have no conflict of interest to declare.

Authors Contribution

MH was responsible for the study concept and design. GB,ZC and MH contributed to the acquisition of data. GBperformed the data analysis. MH assisted with data analy-sis and interpretation of findings. GB drafted themanuscript. MH and ZC provided critical revision ofthe manuscript for important intellectual content. Allauthors critically reviewed content and approved thefinal version for publication.

References

APA (1994) DSM-IV: Diagnostic and Statistical Manual ofMental Disorders. Washington, DC: American PsychiatricAssociation.

Bedi G, Foltin R, Gunderson E, Rabkin J, Hart C, Comer S,Vosburg S, Haney M (2010) Efficacy and tolerability of

high-dose dronabinol maintenance in HIV-positive marijuanasmokers: a controlled laboratory study. Psychopharmacology212:675–686.

Ben Amar M (2006) Cannabinoids in medicine: a review of theirtherapeutic potential. J Ethnopharmacol 105:1–25.

Brook JS, Zhang C, Brook DW (2011) Developmental trajectoriesof marijuana use from adolescence to adulthood. Arch PediatrAdolesc Med 165:55–60.

Budney A, Roffman R, Stephens R, Walker D (2007a) Marijuanadependence and its treatment. Addict Sci Clin Pract 4:4–16.

Budney AJ, Moore BA, Rocha HL, Higgins ST (2006) Clinicaltrial of abstinence-based vouchers and cognitive-behavioraltherapy for cannabis dependence. J Consult Clin Psychol74:307–316.

Budney AJ, Vandrey RG, Hughes J, Moore BA, Bahrenburg B(2007b) Oral delta-9-tetrahydrocannabinol suppresses can-nabis withdrawal symptoms. Drug Alcohol Depend 86:22–29.

Chen K, Kandel DB, Davies M (1997) Relationships between fre-quency and quantity of marijuana use and last year proxydependence among adolescents and adults in the UnitedStates. Drug Alcohol Depend 46:53–67.

Folstein MF, Luria R (1973) Reliability, validity, and clinicalapplication of the Visual Analogue Mood Scale. Psychol Med3:479–486.

Fraser A, Meatherall R (1989) Lack of interference by Nabilonein the EMIT® d.a.u cannabinoid assay, Abbott TDx® cannab-inoid assay, and a sensitive TLC assay for delta-9-THC-carboxylic acid. J Anal Toxicol 13:240.

Grassi MC, Enea D, Ferketich AK, Lu B, Pasquariello S,Nencini P (2011) Effectiveness of varenicline for smokingcessation: a 1-year follow-up study. J Subst Abuse Treat41:64–70.

Haney M (2002) Effects of smoked marijuana in healthy andHIV + marijuana smokers. J Clin Psychopharmacol 42:34S–40S.

Haney M (2007) Opioid antagonism of cannabinoid effects: dif-ferences between marijuana smokers and non-marijuanasmokers. Neuropsychopharmacology 32:1391–1403.

Haney M (2008) Self-administration of cocaine, cannabis andheroin in the human laboratory: benefits and pitfalls. AddictBiol 14:9–21.

Haney M, Hart C, Vosburg S, Comer S, Reed S, Foltin R (2008)Effects of THC and lofexidine in a human laboratory model ofmarijuana withdrawal and relapse. Psychopharmacology197:157–168.

Haney M, Hart C, Vosburg S, Nasser J, Bennett A, Zubaran C,Foltin R (2004) Marijuana withdrawal in humans: effects oforal THC or divalproex. Neuropsychopharmacology 29:158–170.

Haney M, Rabkin J, Gunderson E, Foltin R (2005) Dronabinoland marijuana in HIV(+) marijuana smokers: acute effects oncaloric intake and mood. Psychopharmacology 181:170–178.

Haney M, Ward A, Comer S, Foltin R, Fischman M (1999a)Abstinence symptoms following oral THC administration tohumans. Psychopharmacology 141:385–394.

Haney M, Ward A, Comer S, Foltin R, Fischman M (1999b)Abstinence symptoms following smoked marijuana inhumans. Psychopharmacology 141:395–404.

Hart C, Haney M, Vosburg S, Comer S, Foltin R (2005) Reinforc-ing effects of oral Delta-9-THC in male marijuana smokersin a laboratory choice procedure. Psychopharmacology 181:237–243.

Nabilone in marijuana smokers 9

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology

Hart C, Ilan A, Gevins A, Gunderson E, Role K, Colley J, Foltin R(2010) Neurophysiological and cognitive effects of smokedmarijuana in frequent users. Pharmacol Biochem Behav96:333–341.

Hart C, van Gorp W, Haney M, Foltin R, Fischman M(2001) Effects of acute smoked marijuana on complexcognitive performance. Neuropsychopharmacology 25:757–765.

Kadden RM, Litt MD, Kabela-Cormier E, Petry NM (2007) Absti-nence rates following behavioral treatments for marijuanadependence. Addict Behav 32:1220–1236.

Keppel G (1991) Design and Analysis: A Researcher’s Hand-book, 4th edn. Upper Saddle River, NJ: Prentice Hall.

Lemberger L, Rowe H (1975) Clinical pharmacology ofnabilone, a cannabinol derivative. Clin Pharmacol Ther18:720–726.

Lemberger L, Rubin A, Wolen R, DeSante K, Rowe H, Forney R,Pence P (1982) Pharmacokinetics, metabolism and drug-abuse potential of nabilone. Cancer Treat Rev 9:17–23.

Levin F, Mariani J, Brooks D, Pavlicova M, Cheng W, Nunes E(2011) Dronabinol for the treatment of cannabis dependence:a randomized, double-blind, placebo-controlled trial. DrugAlcohol Depend 116:142–150.

Lile J, Kelly T, Hays L (2010) Substitution profile of the cannab-inoid agonist nabilone in human subjects discriminatingdelta-9-Tetrahydrocannabinol. Clin Neuropharmacol 33:235–242.

Lile J, Kelly T, Hays L (2011) Separate and combined effects ofthe cannabinoid agonists nabilone and delta9-THC in humansdiscriminating delta9-THC. Drug Alcohol Depend 116:86–92.

McGilveray I (2005) Pharmacokinetics of cannabinoids. PainRes Manag 10 (Suppl. A): 15A–22A.

Mendelson J, Mello N (1984) Reinforcing effects of oral delta-9tetrahydrocannabinol, smoked marijuana, and nabilone:influence of previous marijuana use. Psychopharmacology83:351–356.

MTP Research Group (2004) Brief treatments for cannabisdependence: findings from a randomized multisite trial.J Consult Clin Psychol 72:455–466.

Nunes EV, Rothenberg JL, Sullivan MA, Carpenter KM, KleberHD (2006) Behavioral therapy to augment oral naltrexone foropioid dependence: a ceiling on effectiveness? Am J DrugAlcohol Abuse 32:503–517.

Parrott AC, Garnham NJ, Wesnes K, Pincock C (1996) Cigarettesmoking and abstinence: comparative effects upon cognitivetask performance and mood state over 24 hours. Hum Psy-chopharmacol 11:391–400.

SAMSHA (2010) Results from the 2009 National Survey onDrug Use and Health. Rockville, MD: SAMSHA, Office ofApplied Statistics.

Sheskin DJ (2000) Handbook of Parametric and NonparametricStatistical Procedures, 2nd edn. Boca Raton, FL: CRC PressLLC.

Stapleton J, Sutherland G (2011) Treating heavy smokers inprimary care with the nicotine nasal spray: randomizedplacebo-controlled trial. Addiction 106:824–832.

Threlkeld M, Parran T, Adelman C, Grey S, Yu J (2006) Trama-dol versus buprenorpine for the management of acute heroinwithdrawal: a retrospective matched cohort control study. AmJ Addict 15:186–191.

United Nations Office on Drugs and Crime (2007) 2007 WorldDrug Report. Vienna, Austria: UNODC.

Vandrey RG, Budney AJ, Hughes JR, Liguori A (2008) A within-subject comparison of withdrawal symptoms during absti-nence from cannabis, tobacco, and both substances. DrugAlcohol Depend 92:48–54.

Ware M, St Arnaud-Trempe E (2010) The abuse potentialof the synthetic cannabinoid nabilone. Addiction 105:494–503.

Wesnes K, Annas P, Edgar C, Deeprose C, Karlsten R, Philipp A,Kalliomaki J, Segerdahl M (2010) Nabilone produces markedimpairments to cognitive function and changes in subjectivestate in healthy volunteers. J Psychopharmacol 24:1659–1669.

Zobel I, Kech S, van Calker D, Dykierek P, Berger M, Schneibel R,Schramm E (2011) Long-term effect of combined interper-sonal psychotherapy and pharmacotherapy in a randomizedtrial of depressed patients. Acta Psychiatr Scand 123:276–282.

10 Gillinder Bedi et al.

© 2011 The Authors, Addiction Biology © 2011 Society for the Study of Addiction Addiction Biology