Antipsychotics for Primary Alcohol Dependence

Abstract

We sought to meta-analytically assess the utility of antipsychotics in patients with primary alcohol dependence. We searched PubMed, Cochrane Library, and PsycINFO without language restrictions from database inception until December 2012, using the following keywords: (randomized, random, OR randomly) AND (placebo) AND (alcohol dependence) AND (neuroleptic OR antipsychotic OR antidopaminergic OR the names of 34 individual antipsychotics). Included in this study were randomized, placebo-controlled trials of antipsychotics lasting ≥ 2 weeks in patients with primary alcohol dependence and without schizophrenia or bipolar disorder. Two independent evaluators extracted data. Standardized mean difference (SMD), risk ratio (RR), and numbers needed to harm (NNH) ± 95% confidence intervals (CIs) were calculated. Across 13 double-blind studies, 1,593 patients were randomly assigned to one of the following: amisulpride (1 study, n = 37), aripiprazole (2 studies, n = 163), flupenthixol decanoate (1 study, n = 142), olanzapine (2 studies, n = 62), quetiapine (4 studies, n = 174), tiapride (3 studies, n = 212), or placebo (13 studies, n = 803). Neither pooled nor individual antipsychotics outperformed placebo regarding relapse prevention (pooled RR = 1.05 [95% CI, 0.95 to 1.16], P = .38, 9 studies, n = 1,405). Antipsychotics were similar to placebo regarding heavy drinking days (P = .15), craving (P = .82), and first alcohol consumption time (P = .94). Placebo outperformed pooled antipsychotics regarding number or percentage of abstinent days/lack of drinking days (SMD = 0.17 [95% CI, 0.01 to 0.33], P = .04, 5 studies, n = 918), without significant group differences after removal of 1 outlying flupenthixol decanoate study (P = .24). Individually, flupenthixol decanoate (1 study, n = 281) was inferior to placebo regarding abstinence/drinking days (P = .004), whereas aripiprazole (1 study, n = 30) was superior regarding heavy drinking days (P < .00001). Antipsychotics caused greater all-cause discontinuation than placebo (RR = 1.24 [95% CI, 1.07 to 1.45], P = .005, NNH = 14), especially aripiprazole (P = .01) and flupenthixol decanoate (P = .001). Discontinuation due to intolerability was similar between antipsychotics and placebo (P = .12), but aripiprazole's risk was higher (P = .003). Drowsiness/sedation (P < .0001, NNH = 9), increased appetite (P = .02, NNH = 14), and dry mouth (P < .0001, NNH = 7) occurred more frequently with pooled antipsychotics. Except for 1 isolated outcome, the studied antipsychotics did not improve abstinence or reduce drinking or craving in patients with primary alcohol dependence.