REVIEW doi:10.1111/add.

14755

Pharmacotherapy for methamphetamine/amphetamine

use disorder—a systematic review and meta-analysis

Brian Chan1,2, Michele Freeman3 , Karli Kondo3, Chelsea Ayers3, Jessica Montgomery3,
Robin Paynter3 & Devan Kansagara1,3,4
Division of General Internal Medicine and Geriatrics, Oregon Health and Science University, Portland, OR, USA,1 Central City Concern, Portland, OR, USA,2 Evidence-
based Synthesis Program Center, VA Portland Health Care System, Portland, USA3 and Department of Medicine, VA Portland Health Care System, Portland, OR, USA4

ABSTRACT

Aims Addiction to methamphetamine/amphetamine (MA/A) is a major public health problem. Currently there are no
pharmacotherapies for MA/A use disorder that have been approved for use by the US Food and Drug Administration or the
European Medicines Agency. We reviewed the effectiveness of pharmacotherapy for MA/A use disorder to assess the qual-
ity, publication bias and overall strength of the evidence. Methods Systematic review and meta-analysis. We searched
multiple data sources (MEDLINE, PsycINFO and Cochrane Library) to April 2019 for systematic reviews (SRs) and
randomized controlled trials (RCTs). Included studies recruited adults who had MA/A use disorder; sample sizes ranged
from 19 to 229 participants. Outcomes of interest were abstinence, defined as 3 or more consecutive weeks with negative
urine drug screens (UDS); overall use, analyzed as the proportion of MA/A negative UDS specimens; and treatment reten-
tion. One SR of pharmacotherapies for MA/A use disorder and 17 additional RCTs met our inclusion criteria encompassing
17 different drugs (antidepressants, antipsychotics, psychostimulants, anticonvulsants and opioid antagonists). We
combined the findings of trials with comparable interventions and outcome measures in random-effects meta-analyses.
We assessed quality, publication bias and the strength of evidence for each outcome using standardized criteria.
Results There was low-strength evidence from two RCTs that methylphenidate may reduce MA/A use: 6.5 versus
2.8% MA/A-negative UDS in one study (n = 34, P = 0.008) and 23 versus 16% in another study (n = 54, P = 0.047).
Antidepressants as a class had no statistically significant effect on abstinence or retention on the basis of moderate strength
evidence. Studies of anticonvulsants, antipsychotics (aripiprazole), opioid antagonists (naltrexone), varenicline and
atomoxetine provided either low-strength or insufficient evidence of no effect on the outcomes of interest. Many of the
studies had high or unclear risk of bias. Conclusions On the basis of low- to moderate-strength evidence, most
medications evaluated for methamphetamine/amphetamine use disorder have not shown a statistically significant benefit.
However, there is low-strength evidence that methylphenidate may reduce use.

Keywords Amphetamine, methamphetamine, pharmacotherapy, stimulant use disorder, substance use disorder,
systematic review.

Correspondence to: Brian Chan, Oregon Health and Science University, 3181 SW Sam Jackson Park Road L475, Portland OR 97239-3098, USA.
E-mail: chanbri@ohsu.edu
Submitted 20 December 2018; initial review completed 8 March 2019; final version accepted 16 July 2019
This article has been contributed to by US Government employees and their work is in the public domain in the USA.

INTRODUCTION Adverse effects of MA/A include restlessness, insomnia,

hyperthermia and possibly convulsions. Long-term use can
Amphetamine and methamphetamine (MA/A) use disor- lead to addiction, paranoia, mood disturbances, agitation,
der is an emerging problem in world-wide. with major psychosis and cognitive impairment [3,4].
medical, psychiatric, cognitive and social consequences [1]. In the United States, according to the Centers for Dis-
According to the United Nations Office on Drugs and ease Control and Prevention (CDC), among all overdose-
Crime’s (UNODC), 2017 report, MA/A is the second most related deaths in 2017, 14.7% were attributed to
common drug used world-wide (approximately 35 million psychostimulants including MA/A, an increase over prior
past year users), and methamphetamine use is increasing years [5]. MA/A accounted for 51.3 emergency depart-
in North America, Oceania and Asia [2]. ment (ED) visits per 100 000 population in 2011, and ED

© 2019 Society for the Study of Addiction Addiction

2 Brian Chan et al.

visits involving stimulants increased 68% between 2009 at least once per week. Detailed study selection criteria
and 2011 [6]. MA/A use is also associated with behavioral are specified in Supporting information, Appendix B. For
consequences, including aggression and criminality, that outcomes related to abstinence and use, we excluded stud-
indirectly lead to morbidity and mortality [7,8]. ies that relied on self-reported drug use, with the exception
Considering the personal and societal costs of MA/A of studies in previous systematic reviews (Table 1).
use disorder, the need for effective treatment strategies is Each title and abstract in the search was screened for
imperative. Currently there are no medications for treat- inclusion by at least one reviewer using pre-specified selec-
ment of MA/A use disorder that have been approved for tion criteria (Supporting information, Appendix B). We
use by the US Food and Drug Administration or European dual-reviewed an enriched batch of high-relevance ab-
Medicines Agency. Behavioral therapies, including cogni- stracts (18.6% of the total search yield) to ensure reliability.
tive behavioral therapy (CBT) and contingency manage- Two investigators independently reviewed the full text of all
ment (CM) are currently the primary interventions for potentially relevant articles for inclusion. All discordant re-
MA/A use disorder. However, it is unclear whether these sults were resolved through consensus or consultation
interventions have durable effect long term, and access to with a third reviewer.
these therapies may be limited for some patients [9].
This article is part of a larger report [10] commissioned
Data abstraction and quality assessment
by the Veterans Health Administration (VHA) and presents
the results of a systematic review examining the benefits One investigator abstracted details related to study design;
and harms of pharmacological treatments for MA/a use setting; population; intervention and follow-up; co-
disorder in adults; we also examine the benefits and harms interventions; outcomes; and harms. A second investigator
of treatment in special populations including patients with confirmed the abstraction. Our outcomes of interest were
co-occurring opioid use disorder (OUD), and 2) subpopula- sustained abstinence, defined as 3 or more consecutive
tions for whom specific pharmacological treatments may weeks of negative urine drug screens (UDS); overall use,
be more or less beneficial. We conducted assessments of which we analyzed as the proportion of UDS samples that
the methodological quality of individual trials, the likeli- were MA/A-negative; and treatment retention, defined as
hood of publication bias, and the overall strength of the the proportion of randomized patients who completed
evidence. treatment; and adverse effects.
Two reviewers independently assessed the risk of bias
METHODS (ROB) of each RCT using criteria developed by the
Cochrane Collaboration [13] (Supporting information,
Data sources and strategies Appendix C). We report the findings from previous system-
We searched Ovid MEDLINE, OvidPsycINFO and Ovid EBM atic reviews as well as their assessments of study quality at
Reviews Cochrane Database of Systematic Reviews, and face value.
gray literature sources to 12 April 2019 (Supporting infor-
mation, Appendix A). We reviewed the bibliographies of Data synthesis and analysis
relevant articles and contacted experts to identify addi-
tional studies. To identify in-progress or unpublished stud- We qualitatively synthesized the evidence and conducted
ies, we searched ClinicalTrials.gov, OpenTrials, and the random-effects meta-analyses [14] to combine the findings
World Health Organization (WHO) International Clinical of trials with comparable interventions and outcome mea-
Trials Registry Platform (ICTRP). The review protocol sures. We used RevMan version 5.3 [15] to calculate the
was registered to the International Prospective Register of overall relative risk (RR) and 95% confidence interval (CI)
Systematic Reviews (PROSPERO) before we initiated the for each outcome in the active treatment group compared
study (CRD42018085667) [11]. Our methods and with placebo. We assessed statistical heterogeneity among
reporting follow Preferred Reporting Items for Systematic the pooled studies using the I2 statistic [16,17]. For studies
Reviews and Meta-Analyses (PRISMA) guidelines [12]. in which an outcome of interest was collected but not
completely reported, we contacted the authors to request
additional data. We classified the overall strength of the
Study selection
body of evidence (SOE) examining each outcome as high,
We included randomized controlled trials (RCTs) of adults moderate, low or insufficient using a method developed
with MA/A use disorder that compared pharmacother- for the Agency for Healthcare Research and Quality’s
apies (head-to-head) or to placebo or psychotherapy. We (AHRQ) Evidence-based Practice Centers (EPCs) [18]. The
excluded studies and comparisons examining patients with SOE ratings are based on consideration of the quality (in-
comorbid psychotic spectrum or bipolar disorders. We ex- ternal validity) of included studies, directness (of the out-
cluded studies that did not perform drug urinalysis (UDS) comes measured and population studied to those of

© 2019 Society for the Study of Addiction Addiction

 Pharmacotherapy for methamphetamine use disorder 3

Table 1 Key questions and scope parameters.

Key question What are the benefits and harms of pharmacotherapy for Are there subpopulations for whom different forms of
MA/A use disorder (alone, or as an adjunct or follow-up to pharmacotherapy are most/least effective for MA/A use
psychosocial treatment)? disorder?
Population Included: non-pregnant adults with MA/A use disorder Subpopulations may include:
Excluded: subjects with psychotic spectrum disorder, bipolar
Demographic factors
disorder Housing status
Severity
Comorbid mental and substance use disorders (e.g. HIV,
mood and anxiety disorders, ADHD, alcohol use, opioid
use/methadone maintained)
Other clinical conditions
Intervention Included: pharmacotherapies identified as a potential treatment for MA/A use disorder (common adjuncts may be med
management; interpersonal therapy; contingency management (or motivational incentives); CBT (including matrix
therapy, relapse prevention)
Excluded: treatment for temporary psychosis associated with stimulant overdose
Comparators Usual care, placebo, or other interventions (control groups should receive the same adjunctive treatments)
Outcomes • Intermediate/behavioral outcomes
Abstinence (UDS only. Self-report only in addition to UDS) Also of interest when available: longest duration of
abstinence (LDA), and whether patients reach at least 3 consecutive weeks (21 or more days) of abstinence.
MA/A use (quantitative urine levels)
Retention in treatment
• Health and other outcomes
Morbidity/mortality
Quality of life
Legal/employment outcomes
• Harms
Study withdrawal due to AE, and severe AE (as reported in the trials)
Timing Minimum study duration (including follow-up) 4 weeks
Settings • Out-patient
• In-patient
• Incarceration/detention centers, correctional facilities
Study design • Randomized controlled trials
• Systematic reviews

ADHD = attention deficit hyperactivity disorder; AE = adverse event; LDA = longest duration of abstinence; CBT = cognitive behavioral therapy; HIV = human
immunodeficiency virus; MA/A = methamphetamine/amphetamine; UDS = urinalysis.

interest to this review), consistency of evidence across tri- RESULTS

als, precision of summary estimates and reporting or publi-
cation bias [19]. We reviewed a total of 5936 citations and selected 369 for
Although the small number of trials for each medica- full text review. One existing systematic review of 17 stud-
tion precluded quantitative analysis for publication bias, ies and 17 additional RCTs that were not included in the
we assessed publication bias qualitatively by considering previous systematic review met inclusion criteria (Fig. 1).
whether or not it was likely that negative studies were se- Sample sizes among the 34 RCTs ranged from 19 to 229
lectively withheld from publication [20]. We considered patients, with mean enrollment of 90 (SD = 53). Seventeen
factors such as number of positive studies included, review different drugs were studied, including antidepressants, an-
of study sponsorship and searching clinicaltrials.gov to en- tipsychotics, psychostimulants, anticonvulsants and opioid
sure no studies that should have been reported but had antagonists (Table 2).
remained unpublished. Table 3 presents a brief summary of findings, and
We separately examined the evidence in subpopula- Table 4 provides a more detailed summary of the evidence
tions including patients with comorbid OUD, alcohol use for each drug or drug class. The characteristics, quality as-
disorder, attention deficit hyperactivity disorder (ADHD) sessment, and findings of primary studies are provided in
and depression. We also examined whether treatment ef- the Supporting information, Appendices C and D.
fects differed by baseline characteristics such as gender, Only three of the 17 studies we identified had a low risk
HIV status, severity of MA/A use and MA/A-negative of bias quality assessments. Many of the included studies
UDS at randomization. had methodological flaws, including poor outcome

© 2019 Society for the Study of Addiction Addiction

4 Brian Chan et al.

Figure 1 Literature flow diagram

Table 2 Number of systematic reviews and primary trials by drug and drug class.

RCTs not in
SRs previous SRs Drug category Drug

1[21] 3 [22–24] Antidepressants Bupropion, mirtazapine, sertraline

– 2 [25,26] Antipsychotics Aripiprazole
– 2 [29,30] Muscle relaxants/anticonvulsants Topiramate, baclofen, gabapentin
– 4 [31–34] Medications for other substance use disorders; opioid antagonists Naltrexone
1 [35] Medications for other substance use disorders; smoking cessation Varenicline
1 [27] Non-stimulant medications for ADHD Atomoxetine
– 4 [36–39] Other pharmacotherapies Citicoline, ondansetron, PROMETA, riluzole
1[21] – Psychostimulants Dexamphetamine, methylphenidate, modafinil

ADHD = attention deficit hyperactivity disorder; RC = randomized controlled trial; SR = systematic review.

reporting, incomplete allocation methods description and conduct meta-analyses in many cases because reported
small sample sizes. In addition, we found high attrition outcomes used various definitions and time-points,
rates in the majority of studies we reviewed. There was preventing comparison to one another.
marked variation across trials in outcome and treatment There were too few trials for any given drug to conduct
adherence reporting (e.g. self-report, biochemical confir- quantitative estimates of publication bias. However, we felt
mation or not reported). This precluded our ability to that there was a low likelihood of publication bias because:

© 2019 Society for the Study of Addiction Addiction

 Pharmacotherapy for methamphetamine use disorder 5

Table 3 Brief summary of findings.

(1) the body of evidence is largely negative—we did not find We found insufficient evidence for the effectiveness of
a disproportionate number of positive studies; (2) most of antidepressants on reducing MA/A use. Findings across tri-
the published studies were not industry-sponsored; and als were mixed, with no benefit reported by the systematic
(3) we searched clinicaltrials.gov and did not find addi- review [21] (three RCTs, n = 122), a modest but statisti-
tional studies that should have been reported [20]. cally non-significant decrease in use with bupropion
reported by a more recent RCT (n = 151) [22] and a statis-
tically significant reduction with mirtazapine in a small
RCT (n = 60) [24]
Antidepressants: bupropion, mirtazapine and sertraline

A previous systematic review [21] and three additional tri- Antipsychotics: aripiprazole
als [22–24] provided evidence on the use of antidepres-
Our search identified two RCTs of aripiprazole for MA/A use
sants for MA/A use disorder. The systematic review [21]
disorder. In one 12-week, unclear-ROB RCT (n = 90), par-
focused on psychostimulants for MA/A use disorder, but
ticipants received either 20 mg of aripiprazole or placebo;
included six RCTs of bupropion, an aminoketone, which
all participants received once-weekly individual relapse pre-
we classified as an antidepressant. Our literature search
vention therapy [25]. The second trial, a 20-week, high-
identified placebo-controlled trials of three antidepressants:
ROB RCT (n = 53), compared 15 mg of aripiprazole to pla-
bupropion (n = 151) [22], mirtazapine (n = 60) [24] and
cebo, with no concurrent interventions [26]. Both studies
sertraline (n = 229) [23].
contribute to low-strength evidence that aripiprazole does
We found moderate-strength evidence that antidepres-
not reduce MA/A use. The evidence for all other outcomes
sants as a class had no statistically significant effect on the
of interest is insufficient; however, it suggests no benefit of
achievement of sustained abstinence (three RCTs in the
aripiprazole and the possibility of increased harm.
systematic review [21] and one additional RCT [23], com-
bined RR = 0.92, 95% CI = 0.63–1.34; Fig. 2) or study re-
Psychostimulants and other medications used for ADHD
tention (four RCTs in the systematic review [21] and three
(dexamphetamine, methylphenidate, modafinil and
additional RCTs [22–24], combined RR = 0.98, 95%
atomoxetine)
CI = 0.89–1.07; Fig. 3). We found low-strength evidence
of no statistically significant difference in severe adverse There were 11 RCTs included in the systematic review [21]
events from two unclear-ROB RCTs [22,24]. of psychostimulants for the treatment of MA/A use

© 2019 Society for the Study of Addiction Addiction

 6

Table 4 Summary of the evidence on pharmacotherapies for MA/A use disorder, stratified by drug or drug class.

N studies per outcome; ROB

Outcome (n = combined participants) Summary of findings Strength of evidencea Comments

Antidepressants (all combined)

Abstinence 1 SR of 3 RCTs [21] (n = 361) No difference. One SR reported and one additional unclear- Moderate Body of evidence with methodological flaws. Consistent
Brian Chan et al.

1 Unclear-ROB RCT [23] ROB RCT reported similar findings findings across studies
(n = 229)
Use 1 SR of 3 RCTs [21] (n = 122) Mixed findings. One SR included 2 RCTs that found no Insufficient Body of evidence with methodological flaws. Indirectness of
1 Unclear-ROB RCT [22] difference in negative UAs, and one small trial (n = 19) that population

© 2019 Society for the Study of Addiction

(n = 151) favored placebo. One additional RCT found no differences in
1 High-ROB RCT [24] (n = 60) week 12 negative UAs and rate of reduction, but a modest
trend towards improvement favoring bupropion (P = 0.09).
One small trial of MSM found that participants receiving
mirtazapine had more negative UAs, with a larger increase in
the number negative UA participants
Retention 1 SR of 4 RCTs [21] (n = 391) No difference. One SR reported a combined number of Moderate Body of evidence with methodological flaws. Indirectness of
2 Unclear-ROB RCTs [22,23] participants not completing the trial OR of 1.10 (95% population. Consistent findings across studies
(n = 380) CI = 0.73–1.67). Two additional RCTs also found no difference
1 High-ROB RCT [24] (n = 60) in retention, and one RCT reported findings favoring placebo
Harms 1 Unclear-ROB RCT [22] Withdrawal due to AEs: NANo difference. Two RCTs found no No evidence: withdrawal Small body of evidence with methodological flaws
(n = 151) difference between groups in reported severe AEs due to AEsLow: SAEs
1 High-ROB RCT [24] (n = 60)
Antidepressants (aminoketone): bupropion
Abstinence 1 SR of 3 RCTs [21] (n = 361) No difference. One SR reported combined abstinence OR of Low Small body of evidence with methodological flaws. Imprecision
1.12 (95% CI = 0.54–2.33)
Use 1 SR of 3 RCTs [21] (n = 122) No difference. One SR included 2 RCTs that found no difference Low Body of evidence with methodological flaws
1 Unclear-ROB RCT [22] in negative UAs, and one small trial (n = 19) that favored
(n = 151) placebo. In addition, one RCT found no differences in week 12
negative UAs and rate of reduction, but a modest trend towards
improvement favoring bupropion (P = 0.09)
Retention 1 SR of 4 RCTs [21] (n = 391) No difference. One SR reported a combined number of Moderate
1 Unclear-ROB RCT [22] participants not completing the trial OR of 1.10 (95%
(n = 151) CI = 0.73–1.67). One additional RCT also found no difference
in retention
(Continues)

Addiction
 Table 4. (Continued)

N studies per outcome; ROB

Outcome (n = combined participants) Summary of findings Strength of evidencea Comments

Harms 1 Unclear-ROB RCT [22] No difference. There was no difference between groups in No evidence: withdrawal Single multi-site study with methodological flaws. Imprecision
(n = 151) reported severe AEs.Withdrawal due to AEs: NA due to AEsInsufficient:
SAEs
Antidepressants (atypical): mirtazapine
Abstinence NA NA No evidence NA

© 2019 Society for the Study of Addiction

Use 1 High-ROB RCT [24] (n = 60) Favors mirtazapine. At the end of the study participants Insufficient Single study with methodological flaws. Indirectness of study
receiving mirtazapine had more negative UAs, with a larger population. Imprecision
increase in the number negative UA participants
Retention All participants: men who have No difference. There was no difference between study groups Insufficient
Harms sex with men No difference. There was no difference in SAEs between study No evidence: withdrawal
groupsWithdrawal due to AEs: NA due to AEsInsufficient:
SAEs
Antidepressants (SSRI): sertraline
Abstinence 1 Unclear-ROB RCT [23] No difference. For participants receiving sertraline with or Insufficient Single multi-site study with methodological flaws. Imprecision
(n = 229) without CM versus placebo with or without CM, there was a
strong trend favoring placebo (P = 0.052)
Use NA NA No evidence NA
Retention 1 Unclear-ROB RCT [23] Favors placebo. Participants receiving sertraline were retained Insufficient Single multi-site study with methodological flaws. Imprecision
(n = 229) for significantly less time than those receiving placebo. When
collapsed, fewer participants receiving sertraline with or
without CM were retained
Harms NA NA No evidence NA
Atypical antipsychotics: aripiprazole
Abstinence 1 Unclear-ROB RCT [25] No difference. One study found no difference in 2+ week Insufficient Single study with methodological flaws. Imprecision
(n = 90) abstinence
Use 1 Unclear-ROB RCT [25] No difference. Neither study found a positive effect of Low Small body of evidence with methodological flaws. Imprecision
(n = 90) aripiprazole on reducing MA/A use (one study significantly
1 High-ROB RCT [26] (n = 53) favored placebo)
Retention 1 High-ROB RCT [26] (n = 53) No difference. No difference in the number of participants who Insufficient Single study with methodological flaws that was ended early
did not complete the trial due to unfavorable interim results. Imprecision
(Continues)
Pharmacotherapy for methamphetamine use disorder

Addiction
7
 8

Table 4. (Continued)

N studies per outcome; ROB

Outcome (n = combined participants) Summary of findings Strength of evidencea Comments

Harms Withdrawal due to AEs: Favors placebo. One unclear-ROB study found significantly Insufficient Small body of evidence with methodological flaws
1 Unclear-ROB RCT [25] more withdrawals due to AEs than placebo A second small
Brian Chan et al.

(n = 90) high-ROB study found no differenceNo difference. One high-

1 High-ROB RCT [26] (n = 53) ROB found no difference in severe AEs
Severe AEs:

© 2019 Society for the Study of Addiction

1 High-ROB RCT [26] (n = 53)
Anticonvulsants and muscle relaxants: topiramate, baclofen, gabapentin
Abstinence 1 Low-ROB RCT [30] (n = 88) No difference. One study of baclofen versus gabapentin versus Insufficient Small body of evidence, imprecision
placebo reported % 3+ weeks abstinence (44 versus 34.6
versus 40.5%; P = NS)
Use 1 Low-ROB RCT [29] (n = 140) Mixed, favors topiramate. Use was significantly lower in Insufficient Small body of evidence, imprecision; indirectness due to
1 Low-ROB RCT [30] (n = 88) topiramate versus placebo; no difference in UA neg samples in topiramate outcome reported
baclofen versus gabapentin versus placebo (50.3 versus 37.1
versus 37.7%; P = 0.58
Retention 1 Low-ROB RCT [29] (n = 140) No difference. Both studies reported non-sig differences in Insufficient Small body of evidence, imprecision
1 low-ROB RCT [30] (N = 88) treatment retention
Harms 1 Low-ROB RCT [29] (n = 140) No difference. Both studies reported no differences in harms Insufficient Small body of evidence, few outcomes reported
1 low-ROB RCT [30] (n = 88)
Medications for other substance use disorders (opioid antagonist): naltrexone
Abstinence 1 unclear-ROB RCT [33] No difference Insufficient 1 RCT in MSM participants; limited applicability to general
(n = 100) population
Use 1 low-ROB RCTs [31] (n = 80) Mixed findings. No consistent evidence of effect Insufficient Inconsistent results and methodological limitations. Higher
3 unclear-ROB RCTs [32–34] rate of negative UA in 1 low-ROB study, but no difference in 3
(n = 300) unclear-ROB studies
Retention 1 low-ROB RCTs [31] and 3 No difference. Treatment retention naltrexone versus placebo: Low Studies reported inconsistent results and risk of bias
2
unclear-ROB RCTs were RR =1.11, 95% CI = 0.88–1.41, I = 61%
included in MA [32–34]
(n = 380)
Harms 1 low-ROB RCTs [31] (n = 80) No difference. WD due to AEOnly 1 of the 4 studies reported Moderate Studies had low and unclear-ROB
3 unclear-ROB RCTs [32–34] severe AEs (3) [33]
(n = 300)
Medications for other substance use disorders (smoking cessation aid): varenicline
Abstinence NA NA No evidence NA
(Continues)

Addiction
 Table 4. (Continued)

N studies per outcome; ROB

Outcome (n = combined participants) Summary of findings Strength of evidencea Comments

Use 1 Unclear-ROB RCT [35] No difference. % free UA( ), mean (SD): 8.6 (10.1) versus 8.1 Insufficient Small body of evidence with small number of participants and
n = 52 (8.2) unclear-ROB due to procedures altered mid-trial

© 2019 Society for the Study of Addiction

Retention 1 Unclear-ROB RCT [35] No difference. 14/27 (52%) versus 12/25 (48%), P = 0.78 Insufficient
n = 52
Harms 1 Unclear-ROB RCT [35] No difference. Severe AEs: none. Withdrawals due to AEs: 1/27 Insufficient
n = 52 versus 1/25, P = 0.96
Psychostimulants: modafinil, dexamphetamine, methylphenidate
Abstinence 1 SR [21]: 2 RCTs of modafinil No difference. In one SR, 2 RCTs of modafinil reported Low Small body of evidence with methodological flaws
(n = 281) abstinence, with combined OR 0.86 (95% CI = 0.46–1.61)
Use 1 SR [21]: 8 RCTs of modafinil, Mixed findings. Six of 8 RCTs in a SR reported no difference in Class: insufficient. Body of evidence with methodological flaws. Findings in favor
dexamphetamine, and negative UAs.For methylphenidate: 2 high-ROB RCTs reported Methylphenidate: low of methylphenidate (low SOE), but not modafinil and
methylphenidate (n = 611) a positive effect on use, while 2 other high-ROB RCTs found no dexamphetamine
difference; 2 RCTs provided incomplete data
Retention 1 SR of 11 RCTs [21] No difference Number of participants who did not complete the Low Body of evidence with methodological flaws. Participants
(n = 1027) trial, combined OR = 1.11 (95% CI = 0.86–1.44) receiving bupropion are included in the combined OR. Findings
were similar for bupropion
Harms NA NA No evidence NA
Non-stimulant medication for ADHD: atomoxetine
Abstinence NA NA No evidence NA
Use 1 Unclear-ROB RCT [27] No difference. Proportion (95% CI) of UA( ) samples:77% Insufficient Small body of evidence with unclear-ROB. Inadequate dosing
(n = 69) (63–91%) versus 67% (53–81%), P = 0.41 adherence noted for atomoxetine
Retention 1 Unclear-ROB RCT [27] No difference. 91 versus 97%, P = 0.317 Insufficient
(n = 69)
Harms 1 Unclear-ROB RCT [27] No difference. Severe AEs: noneWithdrawals due to AE: none Insufficient
(n = 69)

ADHD = attention deficit hyperactive disorder; AE = adverse event; CI = confidence interval; MD = mean difference; NR = not reported; OR = odds ratio; RCT = randomized control trial; RD = risk difference; ROB = risk of bias; RR = risk ratio;
SAE = severe adverse event; SMD = standard mean difference; SR = systematic review; SOE = strength of evidence. aThe overall quality of evidence for each outcome is based on the consistency, coherence and applicability of the body of evidence,
as well as the internal validity of individual studies. The strength of evidence is classified as follows [18]: High = further research is very unlikely to change our confidence on the estimate of effect. Moderate = further research is likely to have an
Pharmacotherapy for methamphetamine use disorder

important impact on our confidence in the estimate of effect and may change the estimate. Low = further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.

Addiction
9
10 Brian Chan et al.

Figure 2 Abstinence in randomized controlled trials (RCTs) of antidepressants versus placebo for methamphetamine/amphetamine (MA/A) use
disorder [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 3 Retention in randomized controlled trials (RCTs) of antidepressants versus placebo for methamphetamine/amphetamine (MA/A) use
disorder [Colour figure can be viewed at wileyonlinelibrary.com]

disorder [21]. Six of these RCTs (one low, one unclear, four fashion (where all missing urine samples were considered
high-ROB) examined methylphenidate, three RCTs (one positive), the mean proportion of amphetamine-negative
unclear, two high-ROB) examined modafinil and two RCTs urine samples was 6.5% in the methylphenidate group
(one low, one unclear-ROB) examined dexamphetamine. and 2.8% in the placebo group (adjusted odds of
In addition, we identified one unclear-ROB RCT of positive urine sample = 0.46, 95% CI = = 0.26–0.81,
atomoxetine, a non-stimulant medication for ADHD that P = 0.008). In a separate trial of patients who began
was evaluated for MA/A use disorder [27]. treatment while in prison (n = 54), the group treated
Overall, there was low-strength evidence that with methylphenidate had a higher proportion of
psychostimulants as a class have no statistically significant amphetamine-negative urines (23 versus 16%,
effect on sustained abstinence, retention or harms. There P = 0.047) after release from prison (weeks 3–24) [28].
was insufficient evidence for effects on MA/A use due to This trial was also rated high risk of bias, owing to incom-
mixed findings. plete outcome data and other sources of potential bias [21].

Atomoxetine
Methylphenidate
We identified one recently published trial of atomoxetine in
The systematic review of psychostimulants [21] included 69 patients with co-morbid OUD on buprenorphine/
five RCTs (n = 323) that examined methylphenidate and naloxone that found no difference in use, retention, or
found no statistically significant effect on treatment harms outcomes compared to placebo [27]. The risk of bias
retention (combined OR = 1.29, 95% CI = 0.67–2.48; in this study was unclear, therefore the findings provide in-
low-strength evidence) [21]. However, two of the five RCTs sufficient evidence to form conclusions.
reported statistically significant reductions in MA/A use
among subjects receiving methylphenidate (low-strength
Anticonvulsants and muscle relaxants: baclofen,
evidence). In ine trial (n = 34) [26], the mean proportion
gabapentin, topiramate
of amphetamine-negative urine samples at 20 weeks in the
methylphenidate group was 32.7% compared to 18.0% in Our search identified two RCTs that examined anticonvul-
the placebo group. When analyzed in an intent-to-treat sants and muscle relaxants for treatment of MA/A use

© 2019 Society for the Study of Addiction Addiction

 Pharmacotherapy for methamphetamine use disorder 11

disorder. One RCT (17 weeks; low ROB; n = 140) compared Comorbid OUD
200 mg of topiramate to placebo, along with once-weekly
We identified only one RCT that examined naltrexone for
behavioral compliance enhancement treatment (BCET)
MA/A use disorder in patients with comorbid OUD [34].
[29]. The second RCT (16 weeks; low ROB; n = 88) was a
The study was a multi-site trial conducted in Russia that
three-arm trial, comparing 200 mg baclofen, 800 mg
randomized 100 patients to receive naltrexone implant
gabapentin and placebo, along with concurrent thrice-
(Prodetoxon 1000-mg implant) and found no statistically
weekly group relapse prevention therapy [30].
significant effect on MA/A use (40 versus 24%,
Findings from one study of topiramate indicate no
P = 0.09), but better retention compared to placebo (52
statistically significant effect on MA/A use, although more
versus 28%, P = 0.01). In addition, more subjects with
patients on topiramate had a 25%+ reduction in UDS-
the naltrexone implant had heroin-negative UAs at
methamphetamine quantity in weeks 6–12 compared
10 weeks (52 versus 20%, P < 0.001) [34]. These findings
with baseline (low-strength evidence) [29]. There were
provide insufficient evidence to form conclusions, but sug-
no statistically significant differences between baclofen,
gest potential benefit.
gabapentin and placebo on any outcome of interest [30].
The evidence for anticonvulsants and muscle relaxants is
insufficient to form conclusions on any outcome of interest. Smoking cessation aid: varenicline

We identified one recently published study (n = 52) of

Medications used for other substance use disorders: varenicline for MA/A use disorder that found no statisti-
naltrexone and varenicline cally significant effect on abstinence, use or retention
(Table 4). The risk of bias for this study was unclear, thus
Opioid antagonists: naltrexone providing insufficient evidence to form conclusions [35].
Our search identified four RCTs examining naltrexone [a
drug approved by the Food and Drug Administration
Other pharmacotherapies: citicoline, ondansetron,
(FDA) for treating opioid and alcohol use disorders] for
PROMETA and riluzole
the treatment of MA/A use disorder at formulations of
50 mg [31], 380 mg extended-release [32,3] and We found four additional studies that examined other
1000 mg implant [34]. We found low-strength evidence drugs or drug combinations in patients with for MA/A
that naltrexone has no statistically significant effect on use disorder No effects were reported in studies of citicoline
study retention (four RCTs [31–34]; RR = 1.11, 95% [36], ondansetron [37] or PROMETA (a combination of
CI = 0.88–3.12; Fig. 4) [31–34] or overall use (three RCTs flumazenil, gabapentin and hydroxyzine) [38]. A high risk
[31–33], combined RR for amphetamine-negative of bias trial of riluzole found statistically significant reduc-
UDS = 1.05, 95% CI = 0.92–1.18; Fig. 4), and moderate tions in use, a statistically non-significant increase in reten-
evidence of no statistically significant difference in harms tion and an increased risk of harms in the treatment arm;
(four RCTs; n = 380) [31–34]. Evidence related to absti- together, these studies provide insufficient evidence to draw
nence is insufficient to form conclusions. conclusions [39].

Figure 4 Retention and overall use in randomized controlled trials (RCTs) of naltrexone versus placebo for methamphetamine/amphetamine (MA/
A) use disorder [Colour figure can be viewed at wileyonlinelibrary.com]

© 2019 Society for the Study of Addiction Addiction

12 Brian Chan et al.

Subpopulations psychostimulants have no statistically significant effect on

abstinence and retention, although methylphenidate may
The systematic review of psychostimulants [21] and three
more effective than placebo in reducing use. Similarly, al-
additional RCTs [22,25,29] (not included in the systematic
though the evidence for anticonvulsants/muscle relaxants
review) examined subgroup differences in adults with
was insufficient, we found low-strength evidence that
MA/A use disorder: MA/A severity at baseline
topiramate is more effective than placebo for reducing
[21,22,25], methamphetamine-negative UDS at randomi-
MA/A use. In addition, there was low-strength evidence
zation [29], gender [22], comorbid or life-time alcohol use
that naltrexone did not improve treatment retention. There
disorder [29], comorbid ADHD [22], comorbid depression
was low-strength evidence that the antipsychotic
[22] and HIV status [25].
aripiprazole has no statistically significant effect on MA/A
Overall, findings are inconclusive due to methodologi-
use. All findings related to subpopulation differences were
cal issues and a limited number of studies examining each
insufficient.
subpopulation. However, it is possible that bupropion [22],
To date, this is the first report, to our knowledge, sum-
but not aripiprazole [25] or psychostimulants [21], may be
marizing the effectiveness of pharmacotherapies for
more effective in reducing MA/A use in individuals who
MA/A use disorder across drug classes. Because there are
are less severely addicted at baseline [22], and topiramate
currently no FDA-approved medications for this increas-
may be more effective in individuals who produce a nega-
ingly clinically relevant condition, the primary goal of our
tive urine screen at randomization [29]. In addition,
review is to aid clinicians in treatment decisions for this
bupropion may be more effective for males with MA/A
high-risk population. In addition, given the limited existing
use disorder than for females, and there is a possibility that
research, we hope that our findings will provide guidance
some individuals with comorbid depression may experi-
to health services researchers in identifying potential sig-
ence more benefit than placebo [22]. No statistically signif-
nals both for further investigation, as well as drugs or clas-
icant differences were found according to ADHD diagnosis
ses that should no longer be pursued. Although our results
[22], life-time alcohol use disorder [29] or HIV status [25]
largely echo those of the prior systematic reviews examin-
(Table S5 in Supporting information, Appendix D).
ing psychostimulants (including bupropion) [21], our re-
view includes a broad perspective across all classes of
DISCUSSION pharmacotherapies, and provides clinicians and re-
searchers with a more holistic view of how to help patients
We identified one systematic review and 17 additional struggling with this condition.
RCTs (not included in the systematic review) of pharmaco- Our findings have several implications. First, against
therapies for treatment of MA/A use disorder. In general, the background of the opioid epidemic, the number of
the research examining pharmacotherapies for MA/A use deaths involving MA/A and other stimulants is increasing
disorder is limited, and with the exception of studies exam- [5]. This review highlights the urgent need for increased
ining anticonvulsants/muscle relaxants, the risk of bias in research into medications to treat this emerging epidemic.
trials are largely high or unclear. Many of the pharmacotherapies reviewed were initially
There was marked variation across trials in outcome targeted for other indications, reflecting the National Insti-
and treatment adherence reporting (e.g. self-report, bio- tute of Drug Abuse’s ‘repurposing’ strategy to accelerate
chemical confirmation or not reported). This precluded development of medications for addictions [40]. Secondly,
our ability to conduct meta-analyses in many cases be- our review highlights the challenge of pharmacotherapies
cause reported outcomes used various definitions and for stimulant use disorder and the chronic disease of addic-
time-points, preventing comparison to one another. Many tion [41] and the need for novel medications specifically
of the included studies had methodological flaws, including targeted toward the neurobiology of MA/A use disorder.
poor outcome reporting, incomplete allocation methods A multi-pronged strategy of medications and behavioral
description and small sample sizes. We also found a wide therapies may be necessary to see lasting effects on absti-
range of medication dosages used in the trials, leading to nence and use. While individual pharmacotherapies we
concern that the lack of effects seen could be due in part reviewed were not effective, combinations of medications
to under-dosing of medications. In addition, we found high may yield different results; we are aware of at least one on-
attrition rates in the majority of studies. going trial of combination bupropion and extended-release
Many of our findings were insufficient to form strong naltrexone for MA/A use disorder that may change conclu-
conclusions. However, we found moderate-strength evi- sions [42]. Thirdly, we identified only one study examining
dence that antidepressants as a class have no statistically MA/A use disorder in subjects with comorbid OUD. Given
significant effect on abstinence or retention, and low- the increasing overlap in patients using both substances
strength evidence of no statistically significant effect on [43], research guiding the treatment of these co-occurring
harms. We found low-strength evidence that substance use disorders is vital. Finally, our review

© 2019 Society for the Study of Addiction Addiction

 Pharmacotherapy for methamphetamine use disorder 13

highlights research gap in defining and reporting of mean- received or pending, or royalties) that conflict with mate-
ingful outcomes. In the era of harm reduction, perhaps rial presented in the report.
abstinence should not be the primary outcome for pharma-
cotherapy—we looked at use reduction and treatment re- Acknowledgements
tention, but often these were incompletely reported in
The authors wish to thank Dr Ingunn Hansdottir and Dr
journals, and trials may not have been powered with these
Johan Franck for supplying additional details about their
outcomes in mind, increasing the risk of bias and limiting
respective studies [31,32]. The findings and conclusions
the ability to detect differences. Coming to a consensus
in this document are those of the authors, who are respon-
concerning what treatment and recovery for MA/A use
sible for its contents; the findings and conclusions do not
disorder ‘looks like’ from a psychosocial or behavioral per-
necessarily represent the views of the Department of Vet-
spective (i.e. standardized definitions of engagement in
erans Affairs or the United States government. This re-
treatment, adherence to medications and behavioral ther-
search was funded by the Department of Veterans Affairs,
apies) and identifying more clinically important outcomes
Veterans Health Administration, Office of Research and De-
would allow for a better evaluation of the effectiveness of
velopment, Quality Enhancement Research Initiative. Dr
pharmacotherapies in the next generation of trials.
Chan’s time was supported by AHRQ sPCOR K12
(K12HS022981).
Limitations

Our systematic review has several limitations. The scope of References

our review was broad, and we relied on existing SRs when
available. We sought to minimize the disadvantages of 1. Karila L., Weinstein A., Aubin H.-J., Benyamina A., Reynaud
M., Batki S. L. Pharmacological approaches to methamphet-
using existing SRs by only including those that met key
amine dependence: a focused review. Br J Clin Pharmacol
quality criteria; conducting updated searches to identify 2010; 69: 578–92.
more recent trials; and combining data in meta-analysis 2. United Nations Office on Drugs and Crime (UNODC). World
from trials in previous systematic reviews with newer trials Drug Report 2017. ISBN: 978–92–1-148291-1, eISBN:
from our search. Our definition of abstinence (three or 978–92–1-060623-3. United Nations publication, Sales no.
E.17.XI.6. 2019-05-17. UNODC. Available at: http://www.
more consecutive weeks) served as a proxy for sustained
unodc.org/wdr2017/field/Booklet_2_HEALTH.pdf (accessed
abstinence, and the effects of treatment on long-term absti- 17 May 2019) (Archived at http://www.webcitation.org/
nence cannot be directly interpolated. Our search was lim- 78RYGi5WL).
ited to English-language studies; however, the likelihood is 3. Glasner-Edwards S., Mooney L. J. Methamphetamine psycho-
low that the exclusion of non-English language studies sis: epidemiology and management. CNS Drugs 2014; 28:
1115–26.
would alter conclusions [44].
4. McKetin R., Dawe S., Burns R. A., Hides L., Kavanagh D. J.,
Teesson M. et al. The profile of psychiatric symptoms exacer-
CONCLUSIONS bated by methamphetamine use. Drug Alcohol Depend 2016;
161: 104–9.
None of the drug classes studied in patients with MA/A use 5. Kariisa M., Scholl L., Wilson N., Seth P., Hoots B. Drug over-
dose deaths involving cocaine and psychostimulants with
disorder had strong or consistent evidence of benefit on
abuse potential – United States, 2003–2017. Morb Mortal
MA/A use, abstinence or treatment retention. Methylphe- Wkly Rep 2019; 68: 388–95.
nidate and topiramate are promising drugs deserving of 6. Crane E. H. Highlights of the 2011 Drug Abuse Warning
further study. Methodological flaws and inconsistent out- Network (DAWN) Findings on Drug-Related Emergency
come reporting temper these conclusions. We found varia- Department Visits. In: The CBHSQ Report. Rockville, MD: Sub-
stance Abuse and Mental Health Services Administration;
tion in treatment dosing and outcomes reporting across
2013, pp. 1–9.
trials that can be addressed in future studies. Continued re- 7. Darke S., Kaye S., McKetin R., Duflou J. Major physical and
search on subpopulations, including patients with co- psychological harms of methamphetamine use. Drug Alcohol
occurring OUD, is necessary to build the knowledge base. Rev 2008; 27: 253–62.
8. Darke S., Kaye S., Duflou J. Rates, characteristics and circum-
stances of methamphetamine-related death in Australia: a
SR registration national 7-year study. Addiction 2017; 112: 2191–201.
PROSPERO CRD42018085667. 9. Lee N. K., Rawson R. A. A systematic review of cognitive and
behavioural therapies for methamphetamine dependence.
Drug Alcohol Rev 2008; 27: 309–17.
Declaration of interests 10. Chan B, Kondo K, Ayers C, Freeman M, Montgomery J,
Paynter R. et al. Pharmacotherapy for Stimulant Use Disor-
No investigators have any affiliations or financial involve- ders: A Systematic Review of the Evidence. VA ESP Project
ment (e.g. employment, consultancies, honoraria, stock no. 05–225. Washington, DC: Department of Veterans
ownership or options, expert testimony, grants or patents Affairs; 2018.

© 2019 Society for the Study of Addiction Addiction

14 Brian Chan et al.

11. Kansagara D., Kondo K., Freeman M., Ayers C., Paynter R., 27. Schottenfeld R. S., Chawarski M. C., Sofuoglu M., Chooi W. T.,
Montgomery J. et al. Pharmacotherapy for stimulant use dis- Zaharim N. M., M. Yasin M. A. et al. Atomoxetine for
orders: a systematic review. PROSPERO 2018 amphetamine-type stimulant dependence during
CRD42018085667. 2019-05-17. Available at: http://www. buprenorphine treatment: a randomized controlled trial. Drug
crd.york.ac.uk/PROSPERO/display_record.php? Alcohol Depend 2018; 186: 130–7.
ID=CRD42018085667 (accessed 17 May 2019) (Archived 28. Konstenius M., Jayaram-Lindstrom N., Guterstam J., Beck O.,
at http://www.webcitation.org/78RdiyNB0) . Philips B., Franck J. Methylphenidate for attention deficit hy-
12. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA peractivity disorder and drug relapse in criminal offenders
Group. Preferred reporting items for systematic reviews and with substance dependence: a 24-week randomized
meta-analyses: the PRISMA statement. PLOS Med 2009; 6: placebo-controlled trial. Addiction 2014; 109: 440–9.
e1000097. Available at: http://www.prisma-statement.org 29. Elkashef A., Kahn R., Yu E., Iturriaga E., Li S.-H., Anderson A.
(accessed 17 May 2019) (Archived at http://www. et al. Topiramate for the treatment of methamphetamine
webcitation.org/78ReC3Ln8). addiction: a multi-center placebo-controlled trial. Addiction
13. Higgins J., Green S. Cochrane Handbook for Systematic 2012; 107: 1297–306.
Reviews of Interventions Version 5.1.0. 2011. Available at: 30. Heinzerling K. G., Shoptaw S., Peck J. A., Yang X., Liu J., Roll J.
http://handbook.cochrane.org/ (accessed 17 May 2019) et al. Randomized, placebo-controlled trial of baclofen and
(Archived at http://www.webcitation.org/78Rgk6X9K). gabapentin for the treatment of methamphetamine depen-
14. DerSimonian R., Laird N. Meta-analysis in clinical trials. dence. Drug Alcohol Depend 2006; 85: 177–84.
Control Clin Trials 1986; 7: 177–88. 31. Jayaram-Lindstrom N., Hammarberg A., Beck O., Franck J.
15. Review Manager (RevMan) [Computer program]. Version Naltrexone for the treatment of amphetamine dependence:
5.3. Copenhagen: The Nordic Cochrane Centre, The a randomized, placebo-controlled trial. Am J Psychiatry
Cochrane Collaboration, 2014. 2008; 165: 1442–8.
16. Higgins J. P., Thompson S. G. Quantifying heterogeneity in a 32. Runarsdottir V., Hansdottir I., Tyrfingsson T., Einarsson M.,
meta-analysis. Stat Med 2002; 21: 1539–58. Dugosh K., Royer-Malvestuto C. et al. Extended-release inject-
17. Higgins J. P., Thompson S. G., Deeks J. J., Altman D. G. able naltrexone (XR-NTX) with intensive psychosocial
Measuring inconsistency in meta-analyses. BMJ 2003; 327: therapy for amphetamine-dependent persons seeking treat-
557–60. ment: a placebo-controlled trial. J Addict Med 2017; 11:
18. Berkman N., Lohr K., Ansari M., McDonagh M., Balk E., 197–204.
Whitlock E. et al. Grading the Strength of a Body of Evidence 33. Coffin P. O., Santos G.-M., Hern J., Vittinghoff E., Santos D.,
When Assessing Health Care Interventions for the Effective Health Matheson T. et al. Extended-release naltrexone for metham-
Care Program of the Agency for Healthcare Research and Quality: phetamine dependence among men who have sex with
An Update. AHRQ Publication no. 13(14)-EHC130-EF. Rock- men: a randomized placebo-controlled trial. Addiction 2017;
ville, MD: Agency for Healthcare Research and Quality 113: 268–78.
Methods Guide for Comparative Effectiveness Reviews; 2013. 34. Tiihonen J., Krupitsky E., Verbitskaya E., Blokhina E.,
19. Atkins D., Chang S. M., Gartlehner G., Buckley D. I., Whitlock Mamontova O., Föhr J. et al. Naltrexone implant for the treat-
E. P., Berliner E. et al. Assessing applicability when comparing ment of polydrug dependence: a randomized controlled trial.
medical interventions: AHRQ and the effective health care Am J Psychiatry 2012; 169: 531–6.
program. J Clin Epidemiol 2011; 64: 1198–207. 35. Briones M., Shoptaw S., Cook R., Worley M., Swanson A. N.,
20. Guyatt G. H., Oxman A. D., Montori V., Vist G., Kunz R., Moody D. E. et al. Varenicline treatment for methamphet-
Brozek J. et al. GRADE guidelines: 5. Rating the quality of ev- amine dependence: a randomized, double-blind phase II
idence—publication bias. J Clin Epidemiol 2011; 64: 1277–82. clinical trial. Drug Alcohol Depend 2018; 189: 30–6.
21. Bhatt M., Zielinski L., Baker-Beal L., Bhatnagar N., Mouravska 36. Brown E. S., Gabrielson B. A randomized, double-blind,
N., Laplante P. et al. Efficacy and safety of psychostimulants for placebo-controlled trial of citicoline for bipolar and unipolar
amphetamine and methamphetamine use disorders: a sys- depression and methamphetamine dependence. J Affect Disord
tematic review and meta-analysis. Syst Rev 2016; 5: 189. 2012; 143: 257–60.
22. Elkashef A. M., Rawson R. A., Anderson A. L., Li S. H., Holmes 37. Johnson B. A., Ait-Daoud N., Elkashef A. M., Smith E. V., Kahn
T., Smith E. V. et al. Bupropion for the treatment of metham- R., Vocci F. et al. A preliminary randomized, double-blind,
phetamine dependence. Neuropsychopharmacology 2008; 33: placebo-controlled study of the safety and efficacy of
1162–70. ondansetron in the treatment of methamphetamine depen-
23. Shoptaw S., Huber A., Peck J., Yang X., Liu J., Dang J. et al. dence. Int J Neuropsychopharmacol 2008; 11: 1–14.
Randomized, placebo-controlled trial of sertraline and contin- 38. Ling W., Shoptaw S., Hillhouse M., Bholat M. A., Charuvastra
gency management for the treatment of methamphetamine C., Heinzerling K. et al. Double-blind placebo-controlled eval-
dependence. Drug Alcohol Depend 2006; 85: 12–8. uation of the PROMETATM protocol for methamphetamine
24. Colfax G. N., Santos G.-M., Das M., Santos D. M., Matheson T., dependence. Addiction 2012; 107: 361–9.
Gasper J. et al. Mirtazapine to reduce methamphetamine use: 39. Farahzadi M.-H., Moazen-Zadeh E., Razaghi E., Zarrindast M.-
a randomized controlled trial. Arch Gen Psychiatry 2011; 68: R., Bidaki R., Akhondzadeh S. Riluzole for treatment of men
1168–75. with methamphetamine dependence: a randomized, double-
25. Coffin P. O., Santos G.-M., Das M., Santos D. M., Huffaker S., blind, placebo-controlled clinical trial. J Psychopharmacol
Matheson T. et al. Aripiprazole for the treatment of metham- 2019; 33: 305–15.
phetamine dependence: a randomized, double-blind, 40. National Institute on Drug Abuse. 2016–2020 NIDA Strate-
placebo-controlled trial. Addiction 2013; 108: 751–61. gic Plan: Priority Focus Areas. 2019-05-17. Available at:
26. Tiihonen J., Kuoppasalmi K., Fohr J. et al. A comparison of https://www.drugabuse.gov/about-nida/strategic-plan/prior-
aripiprazole, methylphenidate, and placebo for amphetamine ity-focus-areas (accessed 17 May 2019) (Archived at http://
dependence. Am J Psychiatry 2007; 164: 160–2. www.webcitation.org/78RhPEYzj).

© 2019 Society for the Study of Addiction Addiction

 Pharmacotherapy for methamphetamine use disorder 15

41. Volkow N. D., Morales M. The brain on drugs: from reward to Supporting Information
addiction. Cell 2015; 162: 712–25.
42. ClinicalTrials.gov [internet]. Bethesda (MD): National Library Additional supporting information may be found online in
of Medicine (US). 2000 Feb 29. Identifier NCT03078075, the Supporting Information section at the end of the
Accelerated Development of Additive Pharmacotherapy article.
Treatment (ADAPT-2) for Methamphetamine Use Disorder
(ADAPT-2). 2017. Available at: https://clinicaltrials.gov/ct2/
show/NCT03078075 (accessed 17 May 2019) (Archived at Appendix A Search Strategy.
http://www.webcitation.org/78ixc1h3o) (accessed 28 May Appendix B Study Selection Criteria.
2019). Appendix C Quality Assessment.
43. Ellis M. S., Kasper Z. A., Cicero T. J. Twin epidemics: the surg- Appendix D Data Supplement.
ing rise of methamphetamine use in chronic opioid users.
Drug Alcohol Depend 2018; 193: 14–20.
44. Moher D., Pham B., Lawson M. L., Klassen T. P. The inclusion
of reports of randomised trials published in languages other
than English in systematic reviews. Health Technol Assess
2003; 7: 1–90.

© 2019 Society for the Study of Addiction Addiction