Alcohol misuse is responsible for about 4% of all deaths annually, and in the UK it costs the NHS more than £3 billion per year.  A number of medications are currently licensed for the treatment of alcohol dependence. Unfortunately the medications don’t work for everyone and in the USA at least, fewer than 10% of alcoholics receive any drug therapy to help them to stop drinking or at least cut down. So, alternative medications are needed.

The drug gabapentin is licensed for the treatment of epileptic seizures, and it works by modulating GABA (gamma-Aminobutyric acid) activity in the brain. The drug may also be an effective treatment for alcohol dependence. This is because patients report depressed mood, craving, and sleep disturbance as they go through alcohol withdrawal. These symptoms, which can persist for several weeks, are associated with relapse to drinking.  Gabapentin seems to minimise the activation of the stress response in the amygdala that is partially responsible for these effects of alcohol withdrawal.

Previous studies have investigated gabapentin (versus placebo) for the prevention of relapse to drinking in alcohol dependence. All of the previous studies suggested that gabapentin could be effective, although these studies either tended to have small sample sizes, or they gave a dose of gabapentin that was probably too low. The current study from Mason et al (2013) aimed to overcome the limitations of the previous trials, by testing two doses of the drug against placebo in a large sample of alcohol-dependent patients.

Methods

This was a three-arm, double-blind, randomised controlled trial of two doses of gabapentin versus placebo. Alcohol-dependent patients (N = 150), who had been abstinent for alcohol for at least three days, were randomly allocated to groups that received either:

• Placebo
• ‘Low dose’ gabapentin (up to 900mg per day)
• ‘High dose’ gabapentin (up to 1,800mg per day)

The gabapentin groups initially received lower doses of the drug that were gradually increased until they reached the target dose. Participants were then maintained on the target dose for 11 weeks, until in week 12 the dose was titrated downwards such that all participants were receiving placebo at the end of the week. There was also a follow-up visit at week 24, although the report contains conspicuously little detail about that. All participants received psychosocial support in addition to medication, in the form of weekly manual-guided counselling and encouragement to attend local self-help groups (e.g. Alcoholics Anonymous) or other psychological treatments.

Outcome measures were:

• A responder analysis based on the rate of complete abstinence at week 12
• A responder analysis based on the rate of ‘no heavy drinking’ at week 12
• The number of drinks consumed per week over the 12 weeks of the trial
• The number of heavy drinking days per week over the 12 weeks of the trial
• Alcohol craving, sleep quality, and depression over the 12 weeks

Results

• It was notable that the authors altered the primary outcome measures from those that were declared when the trial was registered, although they did justify this, and they reported the originally registered analyses anyway
• 85 of 150 participants completed the study, and the average time that participants remained in the 12-week trial was just over 9 weeks. However, the dropout rates were similar across the different groups

Overall, gabapentin had a clear dose-dependent effect on most of the outcome measures:

• Gabapentin improved the rate of complete abstinence over the course of the 12 week trial, which was:
  • 4.1% in the placebo group
  • 11.1% in the 900mg group
  • 17.0% in the 1,800mg group
  • Number Needed to Treat (NNT) = 8; OR = 4.8, 95% CI = 0.9 to 35 for the 1800mg vs. placebo comparison
• Gabapentin had similar effects on the rate of ‘no heavy drinking’ over the course of the 12 week trial, which were:
  • 22.5% in the placebo group
  • 29.6% in the 900mg group
  • 44.7%  in the 1,800mg group
  • NNT = 5; OR = 2.8, 95% CI = 1.1 to 7.5
• Both doses of gabapentin led to reductions in the number of drinks consumed per week and the number of heavy drinking days per week, over the course of the trial. However, it was not clear if these effects were dose-dependent (the differences between the 900mg and 1800mg doses did not seem to be robust)
• Relative to participants who received placebo, participants who received the high dose (1800mg) showed marked improvements in craving, sleep quality and depressed mood over the course of the trial
• Only 65 participants who completed the 12 week trial returned for the follow-up at week 24.  Nonetheless, the dose-related effects of prior gabapentin treatment on the four drinking-related variables were maintained at follow-up
• Finally, the drug was well tolerated. There were no deaths or serious adverse events. Nine participants dropped out due to adverse events, but these participants were evenly distributed among the placebo, 900mg and 1800mg gabapentin groups

Discussion

The results of this trial are encouraging because they suggest that gabapentin may be an effective medication for the treatment of alcohol dependence, in particular it may reduce the frequency of heavy drinking and increase the rate of abstinence if given during the first few months when people are attempting to abstain from alcohol.

However this trial does have its flaws, not least the decision to switch the primary outcome measures from those that were originally registered, the high dropout rate, and the short-term follow-up period (12 weeks after medication was discontinued), for which data were available from only a small number of participants.

In mitigation, the authors defended their decision to change the outcome measures, and they reported the originally-registered outcome measures anyway; and high dropout rates are unfortunately common for all alcoholism treatments.

One strength of the study is that all patients received a psychosocial intervention as well. We know that even minimal psychosocial interventions are very effective for this population (and we can see this in the present study based on the improvement over time in the placebo group), so any additional benefit of the drug is impressive.

The next step is to investigate how gabapentin (particularly the high dose) compares to medications that are currently licensed for alcohol dependence, such as acamprosate and naltrexone (discussed here by the Mental Elf). One previous trial did test the combination of naltrexone and gabapentin, but did not compare the two medications with each other. It is also notable that although the participants in the present study were alcohol-dependent, none of them required detoxification, which means that they were not severely physically dependent. It is important to find out if the drug is useful in more severely dependent alcoholics. The authors argue that more doctors may be willing to prescribe gabapentin for their alcohol-dependent patients, given that they probably already prescribe the drug for patients with other conditions, and the same cannot be said for acamprosate or naltrexone.

As noted in an enthusiastic commentary on the paper that appeared in the same issue of the journal, the drug is off-patent which means that it would be very cheap to prescribe. Unfortunately this also means that pharmaceutical firms have no financial incentive to get the drug licensed for the treatment of alcohol dependence, and government support would be needed to make this happen.

To end on a note of caution: it is important to identify new treatments for alcohol dependence, including new medications, because the disorder is notoriously difficult to treat and it is desirable for clinicians and patients to have a choice of different treatments. However like all other treatments, gabapentin is unlikely to be a panacea: more than 80% of participants who received the high dose of gabapentin were still drinking at the end of this trial. When you turn the results on their head in this way, the reported odds ratio of 4.8 starts to look a little less impressive.

Links

Mason, B. J., Quello, S., Goodell, V., Shadan, F., Kyle, M., & Begovic, A. (2013). Gabapentin treatment for alcohol dependence: a randomized controlled trial. JAMA Internal Medicine, published online 4^th November 2013, doi: doi:10.1001/jamainternmed.2013.11950

Commentary: Nunes, E. (2013). Gabapentin: A new addition to the armamentarium for alcohol dependence? JAMA Internal Medicine, published online 4^th November 2013.