Titrated baclofen for high-risk alcohol consumption: a randomized placebo-controlled trial in out-patients with 1-year follow-up.

BACKGROUND AND AIMS: Baclofen is a promising drug for treating patients with alcohol-related disorders. Nevertheless, the first randomized clinical trials (mainly with target doses) reported inconsistent efficacy, possibly because of the effective dose widely varying between patients. The Bacloville study aimed to test the efficacy of titrated baclofen for achieving low-risk alcohol consumption. DESIGN: Twelve-month multicenter pragmatic double-blind randomized clinical trial from June 2012 to June 2014. SETTING: Sixty-two French primary care centers. PARTICIPANTS: Out-patients with high-risk alcohol consumption (> 40 g/day for women and > 60 g/day for men). INTERVENTION AND COMPARATOR Patients were randomly assigned (1  1 ratio) to receive titrated baclofen up to 300 mg/day or placebo for 12 months. Switching to open-label baclofen was allowed in cases of perceived inefficacy. MEASUREMENTS The primary outcome defined success as no or low-risk alcohol consumption (≤ 20 g/day for women and ≤ 40 g/day for men) during the last month of the 1-year follow up, with patients who switched to open-label baclofen classified as failures. FINDINGS: A total of 320 patients were randomized, 162 to baclofen and 158 to placebo (consumption 129 g/day in both arms). Discontinuation rates were 30 and 34% in the baclofen and placebo arms, respectively, and return rates of the last-month diaries were 42 and 34%, respectively. Primary success rates were 57 and 36% in the baclofen and placebo arms, respectively [difference 21 percentage points, 95% confidence interval (CI) = 8-34, P = 0.003]. When switchers were not classified as failures unless they failed, the success rates were 62 versus 55% (difference 6 percentage points, 95% CI = -7 to 20). Over 12 months, daily consumption differed between both arms (11 g less in the baclofen arm), as did the number of abstinence days (3.3 days more in the baclofen arm). Adverse events were more frequent with baclofen than placebo and were mostly drowsiness, fatigue and insomnia. Serious adverse events occurred in 85 (seven deaths) and 36 (three deaths) patients with baclofen and placebo, respectively. CONCLUSIONS: Baclofen was more effective than placebo in reducing alcohol consumption to low-risk levels. The number of adverse events and more serious adverse events was greater with baclofen than placebo.