Gueorguieva 2007: a new look at two negative naltrexone trials

What they tested

Two earlier trials, the Veterans Affairs (VA) Cooperative Study #425 and the Women's Naltrexone Study, had failed to show that naltrexone worked on their main measures. This paper went back to their day-by-day drinking records.

For each trial separately, the researchers used a statistical method that sorts people into groups with similar drinking patterns over time, for both any drinking and heavy drinking. They then asked whether naltrexone changed the chance of falling into each group.

What they found

Three drinking patterns showed up in both trials, for any drinking and for heavy drinking. The authors called them abstainer, sporadic drinker and consistent drinker.

Naltrexone doubled the odds of following the abstainer pattern instead of the consistent drinker pattern.

Naltrexone did not significantly change the odds of being in the abstainer pattern compared with the sporadic drinker pattern.

Limits

• It is a reanalysis of trials that were negative on the outcomes they set out to measure. Looking at old data in new ways can turn up findings by chance.
• The abstract doesn't report how many people were in each trial or each drinking pattern.
• The patterns were identified from how people actually drank, after the fact, so they can't be used to pick in advance who will respond.

What it means if you're considering naltrexone

The authors suggested naltrexone may have a meaningful effect for people with a high chance of drinking steadily, even in trials where it failed on the planned analysis. That helps explain why some trials of naltrexone look negative while pooled reviews find a benefit.

It doesn't overturn the two negative results. It's a reason to think the average result in a trial can hide a group of people who do better on the drug.