Mason 2014: gabapentin for alcohol dependence

What they tested

150 adults with alcohol dependence were randomly given gabapentin 900 mg a day, gabapentin 1,800 mg a day or a placebo for 12 weeks. Everyone also had counseling that followed a manual.

The two main measures were rates of complete abstinence and of no heavy drinking over the 12 weeks. The trial also tracked mood, sleep and craving. It ran from 2004 to 2010 at a single outpatient research clinic next to a general hospital.

What they found

Complete abstinence rates were 4.1% on placebo, 11.1% on 900 mg and 17.0% on 1,800 mg. The number needed to treat (NNT), how many people need the drug for one more to benefit, was 8 for 1,800 mg.

Rates of no heavy drinking were 22.5% on placebo, 29.6% on 900 mg and 44.7% on 1,800 mg, with an NNT of 5 for 1,800 mg.

Mood, sleep and craving also improved more as the dose went up.

There were no serious drug-related side effects. Dropouts because of side effects (9 of 150), time in the study and the rate of finishing (85 of 150) did not differ between groups.

Limits

• It was a single-site trial, and only 85 of the 150 people finished it.
• Everyone also had counseling, so the results apply to gabapentin used alongside counseling.
• It had no naltrexone group, so it can't say which drug works better.

What it means if you're considering naltrexone

This trial found clear dose-related benefits for gabapentin, but it was small, ran at one site and lost many participants before the end. The site's guide on medicines to stop drinking notes that the VA/DoD guideline cites the single site and high dropout as reasons for caution, and that a later multisite trial of an extended-release form found no difference in craving.