shearer et al 2009 addiction

8/17/2019 Shearer Et Al 2009 Addiction

1/10

A double-blind, placebo-controlled trial of modafinil

(200 mg/day) for methamphetamine dependence

James Shearer 1, Shane Darke1, Craig Rodgers2, Tim Slade1, Ingrid van Beek2, John Lewis3,

Donna Brady4, Rebecca McKetin1, Richard P. Mattick1 & Alex Wodak4

National Drug and Alcohol Research Centre, University of New South Wales, Sydney, NSW, Australia,1 Kirketon Road Centre, Sydney Hospital, Sydney, NSW,

Australia,2 Pacific Laboratory Medicine Ser vices, Sydney, NSW, Australia3 and St Vincent’s Hospital, Sydney, NSW, Australia4

ABSTRACT

Aim To examine the safety and efficacy of modafinil (200 mg/day) compared to placebo in the treatment of meth-

amphetamine dependence and to examine predictors of post-treatment outcome. Participants and design Eighty

methamphetamine-dependent subjects in Sydney, Australia were allocated randomly to modafinil (200 mg/day)

(n = 38) or placebo (n = 42) under double-blind conditions for 10 weeks with a further 12 weeks post- treatment

follow-up. Measures Comprehensive drug use data (urine specimens and self-report) and other health and psycho-

social data were collected weekly during treatment and research interviews at baseline, week 10 and week 22.

Results Treatment retention and medication adherence were equivalent between groups. There were no differences

in methamphetamine abstinence, craving or severity of dependence. Medication-compliant subjects tended to provide

more methamphetamine-negative urine samples over the 10-week treatment period (P = 0.07). Outcomes were better

for methamphetamine-dependent subjects with no other substance dependence and those who accessed counselling.

There were statistically significant reductions in systolic blood pressure (P = 0.03) and weight gain (P = 0.05) in

modafinil-compliant subjects compared to placebo. There were no medication-related serious adverse events. Adverse

events were generally mild and consistent with known pharmacological effects. Conclusions Modafinil demonstrated

promise in reducing methamphetamine use in selected methamphetamine-dependent patients. The study findings

support definitive trials of modafinil in larger multi-site trials.

Keywords Abuse liability, counselling, methamphetamine dependence, modafinil, randomised controlled trial,safety, treatment.

Correspondence to: Shane Darke, National Drug and Alcohol Research Centre, University of New South Wales, Sydney, NSW 2052, Australia.

E-mail: [email protected]

Submitted 24 May 2008; initial review completed 22 August 2008; final version accepted 15 October 2008

INTRODUCTION

Methamphetamine dependence is a major public healthissue in many parts of the world [1–3] associated with a

wide range of psychological, medical and social problems

[4]. Long-term use of high-dose methamphetamine dys-

regulates monoamine-based neurotransmission, which

is hypothesized to form a neurobiological basis of meth-

amphetamine dependence and withdrawal [5–8]. Dys-

regulation of reward-related glutamate pathways, as

observed in cocaine use, represents a further potential

neurobiological substrate for methamphetamine depen-

dence [7,9,10]. Despite increasing knowledge of the neu-

robiological consequences of methamphetamine use, no

medications to date have been any more successful than

placebo in reducing methamphetamine use in dependent

patients [11,12].Modafinil is a novel non-amphetamine-type stimulant

approved in Australia for thetreatmentof narcolepsy, shift

workers syndrome and sleep apnoea-related fatigue.

Modafinil acts specifically to promote wakefulness and

vigilance, although its precise pharmacological mecha-

nism of action is yet to be delineated [13–16]. The

dopamine- and glutamate-enhancing actions of modafi-

nil have been postulated as beneficial in reducing

withdrawal severity attributable to psychostimulant-

induced neuroadaption [17,18].The stimulant properties

of modafinil may also attenuate the disturbed sleep

RESEARC H REPO RT doi:10.1111/j.1360-0443.2008.02437.x

© 2009 The Authors. Journal compilation © 2009 Society for the Study of Addiction Addiction, 104, 224–233

mailto:[email protected]:[email protected]


2/10

patterns, poor concentration and low energy levels char-

acteristic of methamphetamine withdrawal.

Modafinil is a well-tolerated drug with a low overdose

risk and few side effects [19]. Modafinil does not appear to

be behaviourally reinforcing, as use does not produce

euphoria and there is no discontinuation effect [20,21].

The effects of modafinil cannot be intensified through

high-dose intravenous or intrapulmonary use due to

insolubility in water and instability at high temperatures

[22]. The safety of modafinil in cocaine users has been

investigated in double-blind placebo-controlled interac-

tion studies. Combined cocaine and modafinil (0 mg,

400 mg, 800 mg daily) produced no haemodynamic

interactions and significantly reduced subjective drug

effect ratings for cocaine [23,24].

Cocaine-dependent patients who received modafinil

(400 mg) as a single morning dose provided significantly

more cocaine-negative urine samples compared to place-

bos during an 8-week trial [25]. There were, however, nosignificant differences on self-reported outcomes, includ-

ing frequency of cocaine use, spending on cocaine or the

severity of cocaine withdrawal and craving. No serious

adverse events or overuse of modafinil were reported. The

safety and tolerability of modafinil (400 mg/day) during

a 10-day in-patient methamphetamine withdrawal pro-

gramme was examinedin an open-label comparison with

mirtazapine (60 mg/day) and pericyazine (2.5–10 mg/

day) [21]. Modafinil and mirtazapine were found to be

safe and well tolerated, with no discernable reinforcing

effects. Both were more effective than pericyazine in

reducing withdrawal symptoms, with modafinil achiev-ing less severe withdrawal symptoms and less sleep

disturbance than mirtazapine. Randomized controlled

placebo trials of modafinil in methamphetamine users

were recommended.

The aim of the present study was to examine the

safety and efficacy of modafinil (200 mg/day) in

reducing methamphetamine use and associated harms

through a randomized, double-blind, placebo-controlled

10-week trial of modafinil (200 mg/day) in 80

methamphetamine-dependent subjects. Relapse and pre-

dictors of methamphetamine use other than group (base-

line characteristics, severity of dependence, treatmentexposure) were also examined 12 weeks post-treatment.

METHODS

Subjects

Eighty methamphetamine-dependent subjects were

recruited through a primary health care centre (n = 72)

anda specialist drug andalcohol service(n = 8) located in

the innercity area of Sydney. Subjects met Diagnosticand

Statistical Manual (DSM-IV) criteria for methamphet-

amine dependence operationalized using the Composite

International Diagnostic Interview [26] and were

current, regular (2–3 days of use perweek or more) meth-

amphetamine users confirmed by a methamphetamine-

positive urine spot test at baseline interview. Individuals

who had an active, serious and uncontrolled medical or

psychiatric illness were excluded; however, those with

comorbiddrug dependence or mental illness were enrolled

after consultation with their usual medical attendants.

Women of child-bearing age were excluded if they were

pregnant, nursing infants or unwilling to avoid preg-

nancy through abstinence or barrier contraception. The

latter requirement was necessary, as modafinil induces

metabolism of steroidal contraceptives. Thirty-seven

methamphetamine users who presented for treatment

were excluded due to incomplete enrolment (n = 10),

ineligibility (mainly infrequent use/long-term abstinence

n = 13) or were eligible but declined either randomization

or medication (n = 14). Seven sets of partners wererecruited into the trial and randomized together (modafi-

nil n = 4, placebo n = 3);however, datawereincludedfrom

onlythe presentingindexcase, as theirpartners would not

constitute independent observations. Subjects provided

written informed consent to participate in the trial

approved by three institutional human research ethics

committees, including that of the University of New

South Wales. Subjects were reimbursed $10 for each

research interview and urine specimen provided during

the trial.

Design

Eighty eligible methamphetamine-dependent subjects

seeking treatment were randomized equally to receive

either modafinil (200 mg/day) (n = 38) or placebo

(n = 42) under equivalent conditions for a period of 10

weeks. The sample size of 40 per group was calculated

based on an estimated between-group difference of 40%

in the proportion of stimulant-positive urine samples at

week 10, with missing samples deemed positive and

allowing a 25% attrition rate (a = 0.05, 1-b = 0.9). A

fixed-dose regimen was selected because modafinil has

dose-independent pharmacokinetics between 200 and600 mg/day [19], and early research in cocaine users

found no difference in cocaine suppression or reductions

in subjective effects between 200-mg and 400-mg doses.

Randomization was conducted independently through

Sydney Hospital Pharmacy. Trial medication was dis-

pensed on a weekly basis through Sydney Hospital Phar-

macy using medication event monitoring system (MEMS)

cap bottles to record unsupervised regimen adherence.

All subjects were offered a brief four-session cognitive

behavioural intervention developed specifically for meth-

amphetamine users [27]. Data were collected on alldrug-

Modafinil for methamphetamine dependence 225



3/10

related counselling received by subjects whether within

or outside the trial during the 10-week treatment phase.

Subjects attended weekly medical reviews for side effects,

adverse events, craving, vital observations, renewal of

weekly scripts and to provide study urine samples. Sub-

jects were re-interviewed 12 weeks after they completed

treatment about their post-treatment progress and drug

use. There were no between-group differences in study

follow-up at week 10 (modafinil 84%; placebo 88%) and

week 22 (modafinil 68%; placebo 62%) (see Fig. 1).

Measures

The Opiate Treatment Index (OTI) [28,29] was used tomeasure self-reported methamphetamine use, other drug

use and other treatment outcomes. A ‘cost of habit’

section was added to capture changes in average expen-

diture on methamphetamine. Additional self-reported

methamphetamine use was collected using 28-day drug

use diaries. Prospective drug use diaries are acknowl-

edged as an accurate and valid method of collecting drug

use data [30]. Psychopathology was assessed using the

Brief Symptom Inventory (BSI) [31]. A positive BSI case

was defined as a score above the 90th percentile for US

non-patient norms on the Global Severity Index or in two

or more of the psychiatric domains. Changes in the sever-

ity of dependence were measured using the Severity of

Dependence Scale (SDS) [32]. Subjects rated their stron-

gest methamphetamine craving in the past week on a

100-mm visual analogue scale (VAS) from ‘none’ to

‘strongest ever’. Urine specimens were collected weekly,

whilesubjectsremainedin treatmentand also at research

follow-up interviews. Urine was collected in jars with

heat-sensitive strips to confirm that the samples were at

body temperature and not substituted or otherwise con-

taminated. The specimens were analysed by immunoas-

say for amphetamine and cocaine metabolites using a

cut-off of 300 mg/l. High-performance thin-layer chro-

matography and mass spectrometry were used for

unequivocal confirmation of methamphetamine [33]

indicative of recent use (within the past 48–72 hours). In

order to monitor further adherence with modafinil, all

urine samples were tested for the presence of modafinilic

acid (urinary metabolite of modafinil) using mass

spectrometry.

Data analysis

Data were analysed using SPSS for Windows (version

15.0) and SAS (version 9.1) PROC GENMOD. An

Baseline

MODAFINIL

(n=38)

All commenced

treatment

Week 10

Completed (n=11)

Followed Up

(n=32)

Week 22

Followed Up (n=26)

Analysed (n=38)

Week 10

Completed (n=15)

Followed Up

(n=37)

Week 22

Followed Up (n=26)

Analysed (n=42)

Lost (n=16)

No reason n=8

Moved n=3

Jail n=2

Adverse event n=3

Baseline

PLACEBO

(n=42)

All commenced

treatment

Lost (n=12)

No reason n=7

Moved n=3

Jail n=1

Adverse event n=1

Assessed

(n=124)

Randomised

(n=80)

Excluded (n=44)

Incomplete n=10

Ineligible n=13

Declined n=14

Partnern

=7

Figure 1 Study flow chart

226 James Shearer et al.



4/10

intention-to-treat approach was followed. Data were col-

lected at research interviews conducted at baseline,

week 10 and week 22, regardless of treatment status.

Data for subjects lost to study follow-up were imputed

from baseline using a worst-case scenario assumption of

no change from baseline. Baseline scores and group

variables were entered through the analysis of covari-

ance (ANCOVA) approach as predictors in the relevant

analyses [34]. All statistical tests were two-tailed using

a 0.05 significance level and 95% confidence (CI)intervals; t-tests were used for normally distributed or

transformed continuous variables, the c2 statistic for

categorical variables and the Mann–Whitney U -statistic

for skewed data. Survival analysis based on the log-rank

(Mantel–Cox) statistic was used to examine between-

group differences in retention. Outcome predictors other

than group (comorbidity, other drug dependence, coun-

selling uptake) were examined by linear regression. Cat-

egorical and continuous repeated measures, including

vital signs, craving and urine drug screens collected at

weekly intervals while subjects remained in treatment,

were examined using generalized estimating equations[35,36]. After examination of the within-subject corre-

lation structures, an autoregressive working matrix

(which assumes that pairs of measures are correlated

more highly than measures that are further apart) was

adopted. Missing samples and observations due to varied

scheduling, lost, de-identified or testing failures were

deemed missing completely at random and results were

imputed on a last observation carried forward basis.

Missing data due to treatment dropout were not imputed

in order not to obscure medication effects, whether posi-

tive or negative [37].

RESULTS

Baseline characteristics

Thetypical subject was male, aged mid-30s, andwelledu-

cated (Table 1). Most subjects injected methamphet-

amine on a daily basis, with a third who preferred to

smoke crystalline methamphetamine. There tended to be

more human immunodeficiency virus (HIV)-positive

cases in the modafinil group (all in gay men) and more

subjects in the placebo group were employed at baseline.

Retention

Eleven modafinil-group subjects (29%) and 15 placebo-

group subjects (36%) completed the 10-week course

of treatment. There was no difference in retention over

the 10-week medication phase [log-rank (Mantel–Cox)

c2 = 0.69, P = 0.41] (Fig. 2). Subjects were asked their

reasons for ceasing medication. Similar proportions

between groups reported medication dropout due to no

perceived effect/benefit (modafinil 33%, nine of 27;

placebo 30%, eight of 27), unacceptable side effects

(modafinil 11%, three of 27; placebo 7%, two of 27) or

moving, entering drug rehabilitation or jail (modafinil

11%, three of 27; placebo 33%, seven of 27). Modafinil

subjects tended to report ceasing medication early due to

achieving abstinence (22%, six of 27) compared to

placebo (7%, two of 27) (c2 = 2.7, P = 0.1).

Adherence

Subjects who received modafinil remained on medication

for a mean of 40 days (26 days) and were medication-

adherent on a mean of 31 days (21 days) as confirmed

Table 1 Baseline characteristics.

Modafinil (n = 38) Placebo (n = 42)

Mean age (years) 35.8 6.9

range 22–50

36.1 9.1

range 20–58

t = -1.85, P = 0.85

Gender (male) 63% 62% c2 = 0.13, P = 0.91

Education (completed year 12) 68% 62% c2 = 0.372, P = 0.54

Employed 43% 61% c2 = 2.243, P = 0.17

Years of regular methamphetamine use 6.9 6.3 7.2 9.7 U = 693.5, P = 0.310

Injecting drug use 68% 57% c2 = 1.083, P = 0.30

Current opioid treatment 13% 12% c2 = 0.029, P = 0.87

Alcohol >6 standard drinks daily 3% 10% c2 = 1.617, P = 0.36

HIV-positive (self-report) 38% 18% c2 = 3.506, P = 0.08

HCV-positive (self-report) 38% 26% c2 = 1.144, P = 0.29

Daily or almost daily use 55% 69% c2 = 1.617, P = 0.20

Mean days used (past 28) 18.4 6.8 20.5 6.5 t = 1.371, P = 0.17

Daily cost of methamphetaminea $95 $79 $107 $90 t = -0.694, P = 0.49

HCV = hepatitis C virus; HIV = human immunodeficiency virus. aAnalysis performed after log-transformation of skewed data.




5/10

by MEMS caps and pharmacy records. This represented

78% adherence to daily medication, almost identical to

placebo adherenceof 77%(mean34 21 days adherent

divided by mean 44 21 days retained). Thus, on an

intention-to-treat basis, adherence was 44% in the

modafinil group and 49% in the placebo group. The

majority of unsupervised doses were taken before 3 p.m.

in both groups (modafinil 79%; placebo 84%). Modafinil

(as modafinilic acid) was detected in the urine of all

modafinil-treated subjects and was not found in any of

the placebo group during the medication phase. Medica-tion overuse was measured through the dispensing of 9

days of drug supply for each 7-day period in MEMS con-

tainers and tablet returns for both groups. There was no

difference in the proportion of subjects who over-used

their takeaway medication [modafinil 47%; placebo 51%

(P = 0.8)] or in the average number of additional tablets

consumed [modafinil 13 10, range 2–32; placebo

22 22, range 2–74 (P = 0.5)]. There was no difference

between either the mean total number of internal

counselling sessions attended [modafinil 0.6 1.6;

placebo 0.5 0.9 (P = 0.7)], external counselling

sessions attended [modafinil 1.1 2.4; placebo 1.4 3.3 (P = 0.6)] or the proportions who did not attend any

counselling [modafinil 53%; placebo 47% (P = 0.6)].

Stimulant use

Urinalysis results

The chart in Fig. 3 illustrates the proportion of urine

samples positive for illicit psychostimulants [cocaine,

methamphetamine or 3,4-methylenedioxymethamphe-

tamine (MDMA)] over the 10-week medication period.

There was a trend (P


6/10

most measures and follow-up points (Table 2). Modafinil

subjects provided more stimulant negative weekly urine

samples in the 4 weeks matching the 28-day self-reportat week 10 (modafinil 1.4 1.4; placebo 0.7 0.9,

P = 0.16). When analyses were adjusted post hoc for HIV-

positive subjects who were over-represented in the

modafinil group, week 10 estimates improved in the

direction of treatment [B = 3.59 (95% CI: -0.45, 7.64),

P = 0.08] but did not alter appreciably at week 22. When

the analysis was limited post hoc to single diagnosis,

methamphetamine-dependent subjects with no other

drug dependence (i.e. opioid maintenance patients

excluded) self-reported days of stimulant use at week 10

were 8.3 8.2 (modafinil) and 13.5 10.2 (placebo)

[B = 4.22 (95% CI: 0.08, 8.35), P <

0.05] and at week 22were 11.6 9.2 (modafinil) and 16.4 10.4 (placebo)

[B = 3.82 (95% CI: -0.60, 8.25), P = 0.09].

Study power

A sample size of 58 per group was calculated retros-

pectively as necessary to detect reliably the observed

between-group difference of 20% using a repeated-

measures design without imputation for missing data

(a = 0.05, 1-b = 0.9, T = 10, r = 0.5) [38]. Alternatively,

a sample size of 136 per group would be necessary to

detect the between-group difference in frequency of

stimulantuse at week 10 with missing data imputed from

baseline (a = 0.05, 1-b = 0.8, D = 3.15/9.2 = 0.34). The

effect size estimate (D) was calculated using the adjusted

between group difference in 28-day stimulantuse at week

10 divided by the pooled standard deviation (see Table 2).

The difference in samplesize estimates reflects the greater

statistical power of repeated-measures designs.

Tolerability

There were no medication-related serious adverse events

during the study. The most commonly reported adverse

events are reported in Table 3. Insomnia was the most

common complaint, which did not differ between groups.

Headache was the second most common complaint, anddid not differ between groups except when severity was

taken into account. The modafinil group reported more

transient mild/moderate headache consistent with

product information. Mood problems were unique to the

modafinil group, although all cases were due to an exac-

erbation of pre-existing symptoms. One subject ceased

modafinil due to panicattacks and elevated anxiety, while

another managed anxiety symptoms through dose

reduction to 100 mg/day. There was a statistically signifi-

cant reduction in systolic blood pressure over time in

the modafinil group compared to placebo (c2 = 19.55,

P = 0.03). Differences in weight gain/loss between thegroups were due mainly to acute weight loss (>4 kg)

observed in four subjects in the placebo group.

Other self-report outcomes

There were no apparent between group differences on a

range of secondary outcomes. There was no difference in

weekly craving during the medication phase (F = 1.839,

P = 0.19) or the SDS score at either treatment follow-up

(modafinil 7.3 3.4; placebo 7.3 3.4) or post-treat-

ment follow-up (modafinil 8.1 3.4; placebo 7.7

4.3). Both groups experienced a substantial regression tothe mean in terms of the overall proportion of BSI cases

(modafinil 95% at baseline to 66% at week 22; placebo

90% at baseline to 60% at week 22). Sixty per cent of

modafinil subjects guessed correctly when asked at week

22 if they had received active medication compared to

55% of placebo subjects (c2 = 0.271, P = 0.6).

Predictors of post-treatment outcome other than group

The relationship between potential predictors of post-

treatment outcome (other than group) and self-reported

Table 2 Self-reported stimulant use.

Measure Period Modafinil (n = 38) Placebo (n = 42) Coefficient B (95% CI) P-value

Mean 28-day dr ug use Baseline 18.4 6.8 20.5 6.5

Week 10a 8.8 8.3 13.2 10.1 3.15 (-0.71, 7.01) 0.11

Week 22a 11.8 9.0 15.9 10.7 2.88 (-1.22, 6.99) 0.17

Change score from baseline Week 10 -9.6 8.4 -7.3 9.4 U = 650, 0.15

Week 22 -6.7 8.8 -4.5 9.9 U = 699, 0.20

Mean quantity used (OTI) Baseline 2.3 2.0 2.7 2.3

Week 10a,b 0.6 1.0 1.5 2.2 0.78 (0.03, 1.53) 0.05

Week 22a,b 1.2 1.3 1.8 2.0 0.54 (-0.19, 1.27) 0.20

Mean spending ($/day) Baseline $95 79 $107 90

Week 10a,b $41 70 $64 86 17.5 (-17, 52) 0.24

Week22a,b $57 82 $75 95 13 (-12, 39) 0.23

CI = confidence interval; OTI = Opiate TreatmentIndex. aBaseline score included in analysisof covariance; banalysis performed after log-transformation.




7/10

28-day stimulant use were examined by linear regression

(Table 4). The most pronounced effect was from attend-

ing counselling. Each session of counselling attended

during the 10-week treatment period reduced 28-day

stimulant self-report by 1 day (95% CI: -1.7, -0.3) at

post-treatment follow-up. Simply attending any form of

counselling reduced 28-day self-report by6 days (95% CI:

-10.8, -1.8). HIV-positive subjects who were over-

represented in the modafinil group had poorer metham-

phetamine use outcomes. BSI cases also had poorer

methamphetamine use outcomes; however, there werevery few non-cases and these were divided equally

between the groups [modafinil two subjects (5%); placebo

four subjects (10%)]. Opioid maintenance therapy

(methadone or buprenorphine) had no significant direct

effect on outcome and the small number of opioid main-

tenance patients were distributed equally between the

groups (five in each group). Overall, however, the lack of

difference at week 22 between opioid-dependent patients

in the modafinil and placebo groups appeared to mask

greater between-group effect size for single-diagnosis

methamphetamine users (see Self-reported stimulant

use). Other baseline and treatment characteristics did not

appear to influence post-treatment outcome.

DISCUSSION

In this treatment population modafinil showed

promise in reducing methamphetamine use in selected

methamphetamine-dependent patients. However, there

were no differences between modafinil and placebo in

retention, methamphetamine abstinence, methamphet-

amine craving or severity of dependence. Subjects whoremained on medication tended to provide more illicit

psychostimulant negative urine samples than those who

received placebo over the 10-week treatment period. This

was consistent with greater declines in the modafinil

group in self-reported 28-day stimulant use and the

quantity of stimulants used. The between-group differ-

ences in favour of modafinil achieved at week 10 per-

sisted at week 22, consistent with a group-related

treatment effect. Emergent treatment effects may have

been masked by over-representation of HIV-positive sub-

jects in the modafinil group, who had poorer metham-

Table 3 Adverse events and vital signs in medication-compliant subjects.

Event Severity/frequency Modafinil Placebo (P)

Insomnia All reports 12/38 (32%) 17/42 (40%) c2 = 0.683, P = 0.41

Headache Mild/moderate 7/38 (18%) 2/42 (5%) c2 = 3.728, P < 0.05

Severe 2/38 (5%) 4/42 (10%) c2 = 0.522, P = 0.47

Nausea Mild/moderate 2/38 (5%) 5/42 (12%) c2 = 1.102, P = 0.29

Severe 3/38 (11%) 1/42 (2%) c2 = 1.277 P = 0.26

Gastric pain/discomfort All reports 5/38 (13%) 3/42 (7%) c2 = 0.802, P = 0.37

Thought disorder All reports 2/38 (5%) – Nc

Anxiety Severe 3/38 (8%) – Nc

Fatigue All reports – 4/42 (10%) Nc

High blood pressurea Baseline 9/38 (24%) 10/42 (24%) c2 = 0.234, P = 0.63

High blood pressurea Final observation 5/34 (15%) 5/39 (13%) c2 = 0.054, P = 0.82

Weight loss Between baseline and 10 weeks 2/11 (18%) 8/15 (53%) c2 = 3.331, P = 0.07

Weight change (mean kg) Between baseline and 10 weeks 2.7 2.8 0.2 3.4 U = 66, P < 0.05

Nc = Not calculated. aHigh blood pressure was defined as either a systolic reading of 155 mmHg or higher, or a diastolic reading of 90 mmHg or higher.

Table 4 Predictors of stimulant use 12 weeks post-treatment.

Variable Rationale Coefficient B (95% CI) P-value

No. of counselling sessions Treatment exposure -0.99 (-1.73, -0.26)


8/10

phetamine outcomes, and the enrolment of subjects in

concurrent opioid maintenance treatment. Methamphe-

tamine-dependent subjects with no other substance

dependence who received modafinil reported the greatest

reductions in methamphetamine use compared to

placebo.

Modafinil appeared to be safe and well tolerated. There

were no medication-related serious adverse events. The

modafinil group reported more transient mild/moderate

headaches and also tended to report more nausea/gastric

discomfort, all of which resolved after the first week of

treatment. Reports of thought disorder and anxiety were

unique to the modafinil group and indicate caution in

patients with pre-existing psychotic symptoms or anxiety.

Modafinil did not appear to be associated with any eleva-

tion of blood pressure or heart rate. The clinical signifi-

cance of a reduction in systolic blood pressure over the

10-week treatment period in medication-adherent sub-

jects receiving modafinil should not be underestimatedgiven the cardiotoxicity of methamphetamine [4]. There

was no evidence of excessive use of modafinil, consistent

with previous reports of low abuse liability [20]. Signifi-

cant weight gain in the modafinil group could be inter-

preted as a positive treatment outcome, as modafinil

usually decreases food intake [39]. The measurement of

body mass index (BMI) should be considered in future

studies. Finally, reports of fatigue were unique to the

placebo group and were consistent with the wake-

promoting properties of modafinil.

More than one-third of subjects who received modafi-

nil experienced no discernable effects on sleep or atten-tion. The expectation of a rapid, strong stimulating effect

may have contributed to disappointment and early treat-

ment dropout for those who believed they were receiving

placebo. Another potential factor has been identified in

sleep studies, where some patients with previous amphet-

amine treatment or illicit methamphetamine histories

appeared to be less sensitive to the stimulating effects of

modafinil [40–42]. Motivational enhancement therapy

may improve adherence and treatment retention in

future studies, particularly in those where the effect of

modafinil was weak or diminishing. The relatively low

dose used in this study may not have been optimal.Higher doses in the usual range (up to 200 mg b.i.d.)

could improve adherence and treatment retention.

Attendance at any form of counselling although

self-selected, was significantly predictive of better post-

treatment outcome. Although uptake of the specific

cognitive–behavioural treatment (CBT) programme

offered to subjects was minimal, many had pre-existing

counselling arrangements and study participants

accessed a diverse array of counselling services during

the course of the trial. The psychosocial platforms on

which medications targeted toward psychostimulant use

are delivered are critical to their success, and these may

vary according to the needs of target populations [43].

Potential strategies to improve the uptake of counselling

could include contingent incentives to attend counsel-

ling, such as travel reimbursement. Narrative therapy

has also shown recent promise in methamphetamine

users [44].

There was no group difference in methamphetamine

craving. These data suggest that modafinil may not

achieve its effects as an anti-craving agent. This finding

wasconsistent with the findings of Dackis and colleagues,

who found no reduction in craving in cocaine users mea-

sured using a similar visual analogue scale [25]. Lack of

effect on craving may simply be evidence for lack of effect

on methamphetamine use; however, the analyses that

found downward trends in weekly methamphetamine-

negative urine samples and self-reported frequency of

methamphetamine were not matched by any changes in

self-reported craving. It may be more likely that modafinilworks as a drug effect agonist and improves cognitive

functioning, rather than as an anti-craving agent. Emerg-

ing research is drawing attention to the importance of

cognitive impairment, impulsivity and stress as critical

features of problematic psychostimulant use [45,46].

Chronic psychostimulant use reduces activity in areas

of the brain associated with cognitive functioning,

regions where modafinil appears to achieve its wake and

vigilance-promoting effects. Through restoring cognitive

functioning, modafinil may help patients to deal with the

life stressors, cravings, cues and compulsive/obsessive

drug-seeking behaviour which are common features of relapse among psychostimulant users.

The main limitations affecting interpretation of the

results of this study were the absence of an objective

quantitative measure, reliance on self-reported outcomes

and a sample size which was too small to detect reliably

the small differences between modafinil and placebo.

Between-group differences based on self-report were not

supported by urinalysis of specimens collected at the

week 10 and week 22 research follow-up interviews.

While prospective self-report is considered valid and reli-

able in drug and alcohol research, there may be a risk of

differential reporting between groups. The substantialplacebo effect of 45% and the high proportion of modafi-

nilsubjects with no discernable medication effect of 40%,

however, mitigated against potential systematic bias in

self-report in favour of modafinil. The weekly urine drug

screens captured changes in drug use over the 10-week

treatment period time, albeit in compliant subjects.

Increased dose and motivational enhancement therapy

informed by the experience of this trial could improve

medication adherence and avoid the treatment disap-

pointment that may have contributed to selective treat-

ment attrition. Alternative biological measures more




9/10

sensitive or appropriate to longer-term quantitative

changes could be considered, such as hair analysis. Given

the predominance of behavioural and psychosocial

factors in methamphetamine dependence, it may be

unrealistic to expect medication alone to achieve large

treatment effect sizes.

Modafinil appeared to be safe, non-reinforcing and

moderately beneficial in reducing methamphetamine use

in this treatment population. These findings need to be

confirmed by larger multi-centre trials. Benefits were

selective for medication-adherent patients who engaged

in any form of psychosocial therapy or support. Future

studies could benefit from the experience in this trial by

encouraging completion of the full course of medication

and engagement in psychosocial therapies. Future trials

could also examine thevalueof longer maintenance regi-

mens or titration to the higher standard dose of 200 mg

b.i.d.

Clinical trial registration

NCT00123370 Clinicaltrials.gov.

Declarations of interest

None.

Acknowledgements

We thank the staff and clients of the Kirketon Road

Centre, Rankin Court and Sydney Hospital Pharmacy for

their support. Funding was provided by the Australian

Government Department of Health and Ageing. Match-

ing active and placebo trial medication was supplied as a

grant by Cephalon Inc.

References

1. McKetin R., Kozel N., Douglas J., Ali R., Vicknasingam B.,

Lund J. et al. The rise of methamphetamine in Southeast

and East Asia. Drug Alcohol Rev 2008; 27: 220.

2. Maxwell J., Rutowski B. The prevalence of methamphet-

amine and amphetamine abuse in North America: a review

of the indicators. Drug Alcohol Rev 2008; 27: 229.

3. Degenhardt L., Roxburgh A., Black E., Bruno R., Campbell

G., Kinner S. et al. The epidemiology of methamphetamineuse andharm in Australia. Drug Alcohol Rev 2008; 27: 243.

4. Darke S., Kaye S., McKetin R., Duflou J. Major physical and

psychological harms of methamphetamine use. Drug

Alcohol Rev 2008; 27: 253.

5. Everitt B. J., Wolf M. E. Psychomotor stimulant addiction: a

neural systems perspective. J Neurosci 2002; 22: 3312–20.

6. Volkow N. D.,Chang L., Wang G.-J., Fowler J.S., Leonido-Yee

M., Franceschi D. et al. Associationof dopamine transporter

reduction with psychomotor impairment in methamphet-

amine abusers. Am J Psychiatry 2001; 158: 377–82.

7. Spanagel R., Weiss F. The dopamine hypothesis of reward:

past and current status. Trends Neurosci 1999; 22: 521–

7.

8. White F., Kalivas P. Neuroadaptions involved in amphet-

amine and cocaine addiction. Drug Alcohol Depend 1998;

51: 141–53.

9. Trudeau L.-E. Glutamate co-transmission as an emerging

concept in monoamine neuron function. J Psychiatry

Neurosci 2004; 29: 296–310.

10. Kalivas P., McFarland K., Bowers S., Szumlinski K., Xi Z.,

Baker D. Glutamate transmission and addiction to cocaine.

Ann N Y Acad Sci 2003; 1003: 169–75.

11. Shearer J., Gowing L. Pharmacotherapies for problematic

psychostimulant use: a review of current research. Drug

Alcohol Rev 2004; 23: 203–11.

12. Srisurapanont M., Jarusuraisin N., Kittirattanapaiboon P.

Treatment for Amphetamine Dependence and Abuse (Cochrane

Review). Oxford: The Cochrane Library, Update Software;

2001.

13. Murillo-Rodriguez E., Haro R., Palomero-Rivero M., Millan-

Aldaco D., Drucker-Colin R. Modafinil enhances extracellu-

lar levels of dopamine in the nucleus accumbens and

increases wakefulness in rats. Behav Brain Res 2007; 176:

353–7.

14. Madras B. K., Xie Z., Lin Z., Jassen A., Panas H., Lynch L.

et al. Modafinil occupies dopamine and norepinephrinetransporters in vivo and modulates the transporters and

trace amine activity in vitro. J Pharmacol Exp Ther 2006;

319: 561–9.

15. Willie J., Renthal W., Chemelli R., Miller M., Scammell T.,

Yanagisawa M. et al. Modafinil more effectively induces

wakefulness in norexin-null mice than wild-type litter-

mates. Neuroscience 2005; 130: 983–95.

16. Scammell T., Estabrooke I., McCarthy M., Chemelli R.,

Yanagisawa M., Miller M. et al. Hypothalamic arousal

regions are activated during modafinil-induced wakeful-

ness. J Neurosci 2000; 20: 8620–8.

17. Dackis C. New treatments for cocaine abuse. Drug Discov

Today 2005; 2: 79–86.

18. Stoops W., Lile J., Fillmore M., Glaser P., Rush C. Reinforcingeffects of modafinil: influence of dose and behavioural

demands following drug administration. Psychopharmacol-

ogy 2005; 182: 186–93.

19. Robertson P., Hellriegel E. Clinical pharmacokinetic profile

of modafinil. Clin Pharmacokinet 2003; 42: 123–37.

20. O’Brien C., Dackis C., Kampman K. Does modafinil produce

euphoria? Am J Psychiatry 2006; 163: 6.

21. McGregor C., Srisurapanont M., Mitchell A., Wickes W.,

White J. Symptoms and sleep patterns during inpatient

treatment of methamphetamine withdrawal: a comparison

of mirtazapine and modafinil with treatment as usual. J

Subst Abuse Treat 2008; 35: 334–42.

22. Jasinski D.,Kovacevic-Ristanovic R. Evaluation of the abuse

liability of modafinil and other drugs for excessive daytimesleepiness associated with narcolepsy. Clin Neuropharmacol

2000; 23: 149–56.

23. Dackis C., Lynch K., Yu E., Samaha F., Kampman K.,

Cornish J. et al. Modafinil and cocaine: a double-blind,

placebo-controlled drug interaction study. Drug Alcohol

Depend 2003; 70: 29–37.

24. Malcolm R., Swayngim K., Donovan J., Devane C., Elkashef

A., Chiang N. et al. Modafinil and cocaine interactions. Am J

Drug Alcohol Abuse 2006; 32: 577–87.

25. Dackis C., Kampman K., Lynch K., Pettinati H., O’Brien C.

A double-blind, placebo-controlled trial of modafinil for

cocaine dependence. Neuropsychopharmacology 2005; 30:

205–11.

232 James Shearer et al.



10/10

26. WHO. Composite International Diagnostic Interview. Sydney:

WHO Collaborating Centre for Mental Health and Sub-

stance Abuse; 1997.

27. Baker A., Lee N., Claire M., Lewin T., Grant T., Pohlman S.

et al. Brief cognitive behavioural interventions for regular

amphetamine users: a step in the right direction. Addiction

2005; 100: 367–78.

28. Darke S., Hall W., Wodak A., Heather N., Ward J. Develop-

ment and validation of a multi-dimensional instrument for

assessing outcome of treatment among opiate users: the

Opiate Treatment Index. Br J Addict 1992; 87: 733–42.

29. Barrowcliff A., Champney-Smith J., McBride A. Use of a

modified version of the Opiate Treatment Index with

amphetamine users: validation and pilot evaluation of a

prescribing service. J Subst Use 1999; 14: 98–103.

30. Del Boca F., Darkes J. Enhacing the validity and utility of

randomized clinical trials in addictions treatment research:

II. Participant samples and assessment. Addiction 2007;

102: 1194–203.

31. Derogatis L. Brief Symptom Inventory (BSI): Administration,

Scoring and Procedures Manual. Minneapolis: National Com-

puter Systems; 1993.

32. Gossop M., Darke S., Griffiths P., Hando J., Powis B., Hall W.et al. The Severity of Dependence Scale (SDS): psychometric

properties of the SDS in English and Australian samples of

heroin, cocaine and amphetamine users. Addiction 1995;

90: 607–14.

33. Lewis J. Working to recognised standards: a prerequisite for

drug testing in Australia. Ther Drug Monit 2002; 24: 182–

6.

34. Vickers A., AltmanD. Analysing controlled trials with base-

line and follow up measurements. BMJ 2001; 323: 1123–

4.

35. Zeger S., Liang K.-Y. Longitudinal data analysis for discrete

and continuous outcomes. Biometrics 1986; 42: 121–

30.

36. Zeger S., Liang K.-Y., Albert P. Models for longitudinal data:

a generalized estimating equation approach. Biometrics

1988; 44: 1049–60.

37. Mazumdar S., Tang G., Houck P., Dew M., Begley A., Scott J.

et al. Statistical analysis of longitudinal psychiatric data

with dropouts. J Psychiatr Res 2007; 41: 1032–41.

38. Twisk J. Applied Longitudinal Data Analysis for Epidemiology.

Cambridge: Cambridge University Press; 2003.

39. Mackris A., Rush C., Frederich R., Kelly T. Wake-promoting

agents with different mechanisms of action: comparison of

effects of modafinil and amphetamine on food intake and

cardiovascular activity. Appetite 2004; 42: 185–95.

40. Broughton R. J., Fleming J. A. E., George C. F. P., Hill J. D.,

KrygerM. H.,Moldofsky H. et al. Randomized, double-blind,

placebo-controlled crossover trial of modafinil in the treat-

ment of excessive daytime sleepiness in narcolepsy. Neurol-

ogy 1997; 49: 444–51.

41. Guilleminault C., Aftab F., Karadeniz D., Philip P., Leger D.

Problems associated with switch to modafinil—a novel

alerting agent in narcolepsy. Eur J Neurol 2000; 7: 381–4.

42. Girish M., Shephard J., Silber M. Change from conventional

stimulants to modafinil in patients with narcolepsy or idio-

pathic hypersomnia. Sleep 2000; 23: A302.

43. Schmitz J., Mooney M., Moeller F., Stotts A., Green C.,Grabowski J. et al. Levodopa pharmacotherapy for cocaine

dependence: choosing the optimal behavioral platform.

Drug Alcohol Depend 2008; 94: 142–50.

44. Cruikshank C., Montebello M., Dyer K., Quigley A.,

Blaszczyk J., Tomkins S. et al. A placebo-controlled trial of

mirtazapine for the management of methamphetamine

withdrawal. Drug Alcohol Rev 2008; 27: 326–33.

45. Sinha R., Garcia M., Paliwal P., Kreek M. J., Rounsaville B. J.

Stress-induced cocaine craving and hypothalamic–

pituitary–adrenal responses are predictive of cocaine

relapse outcomes. Arch Gen Psychiatry 2006; 63: 324–31.

46. Volkow N., Fowler J. Addiction, a disease of complusion and

drive: involvement of the orbitofrontal cortex. Cereb Cortex

2000; 10: 318–25.



shearer et al 2009 addiction

Documents