Table 2.

GABA agents

Pharmacological agents	Study design Sample characteristics	Outcomes	Results	Main limitations	References
Baclofen (60 mg day−1)	16-week, randomized, placebo-controlled, double-blind trial with 3 arms including gabapentin and placebo as well as baclofen 25 methamphetamine dependent patients Primary route of use: smoking	Primary outcomeMethamphetamine use (urine drug tests) Secondary outcomesTreatment retention, depressive symptoms, methamphetamine cravings, adverse events	Not effective in reducing methamphetamine use Small treatment effect relative to placebo among participants who reported taking a higher percentage of study medication (post hoc analyses)	Small sample size Increasing attrition of the participants biological measures of medication adherence not performed	[80]
Gabapentin (2400 mg day−1)	16-week, randomized, placebo-controlled, double-blind trial with 3 arms including baclofen and placebo as well as gabapentin 26 methamphetamine dependent patients Primary route of use: smoking	Primary outcomeMethamphetamine use (urine drug tests) Secondary outcomesTreatment retention, depressive symptoms, methamphetamine cravings, adverse events	Not effective in reducing methamphetamine use	Small sample size Increasing attrition of the participants biological measures of medication adherence not performed	[80]
Vigabatrin (from 1 to 3 g day−1)	9-week open-label safety study 28 methamphetamine and/or cocaine dependent patients	Primary outcomesChange of visual fields, visual acuity, ocular adverse effects Secondary outcomesMethamphetamine use	No ophthalmologic adverse effects Decrease of methamphetamine use	Small sample size Open label design No control group High drop-out rate (approximately 50%)	[83]
Vigabatrin (from 1.5 to 3 g day−1)	9-week open-label safety study 30 methamphetamine and/or cocaine dependent patients	Primary outcomesChange of visual fields, visual acuity, ocular adverse effects Secondary outcomesMethamphetamine use	No ophthalmologic adverse effects Decrease of methamphetamine use	Small sample size Open label design No control group High drop-out rate (approximately 50%)	[83]