Antipsychotics for cocaine or psychostimulant dependence: systematic review and meta-analysis of randomized,...

Since cocaine and psychostimulant dependence are related to increased dopamine release, antipsychotics have been tried to reduce their reinforcing properties. A meta-analysis was undertaken to assess the efficacy and tolerability of antipsychotics in cocaine- or stimulant-dependent patients.

We searched PubMed, Cochrane Library databases, and PsycINFO from database inception until June 24, 2013, using the following keywords: (randomized OR random OR randomly) AND (placebo) AND (methylphenidate OR cocaine OR methamphetamine OR amphetamine OR 3,4-methylenedioxymethamphetamine) AND (dependence OR abuse) AND (antipsychotic OR neuroleptic OR 34 specific antipsychotic names).

Included were randomized, placebo-controlled trials of antipsychotics lasting at least 2 weeks in patients with primary cocaine or psychostimulant dependence. Of 363 hits, we removed 316 duplicates, 20 references based on abstract/title, and 13 ineligible full-text articles, retaining 14 trials for this meta-analysis.

Two authors independently extracted the data. Coprimary outcomes included degree of substance use and lack of abstinence. Risk ratio (RR), 95% CI, and standardized mean difference were calculated.

Ten studies in patients with primary cocaine dependence (risperidone = 5, olanzapine = 3, reserpine = 2; n = 562) and 4 in those with amphetamine/methamphetamine dependence (aripiprazole = 4; n = 179) were meta-analyzed (14 studies, total n = 741). When study results were pooled together, antipsychotics did not differ from placebo in regard to cocaine use days and lack of cocaine or amphetamine/methamphetamine abstinence, severity of addiction, cocaine or amphetamine/methamphetamine craving, Clinical Global Impressions-Severity of Illness (CGI-S) scores, depression, anxiety, compliance, all-cause discontinuation, and several side effects. However, antipsychotics caused more intolerability-related discontinuation than placebo (P = .0009). Individually, aripiprazole was superior to placebo in regard to CGI-S (P = .001), while olanzapine was inferior to placebo in regard to cocaine craving (P = .03) and risperidone was inferior to placebo in regard to depression (P = .002).

Antipsychotics had no advantages over placebo in regard to cocaine use and cocaine or psychostimulant abstinence or craving, while causing more intolerability-related discontinuations.