Bolton 2019: serious side effects of oral naltrexone across 89 trials

What they tested

The researchers searched for every placebo-controlled randomized trial of oral naltrexone longer than 4 weeks and published after 1 January 2001, for any condition and at any dose. They left out studies in current or former opioid users, because naltrexone and opioids can interact.

They found 89 trials with 11,194 participants, studying alcohol use disorders (38 trials), other addictions including smoking (18), psychiatric disorders (13), impulse control disorders (9), obesity or eating disorders (6), Crohn's disease (2), cancers (2) and fibromyalgia (1). Twenty-six trials, with 4,960 participants, recorded serious adverse events by study group and went into the main analysis.

What they found

In the 26 trials (4,960 participants) that recorded serious adverse events by group, naltrexone didn't raise the risk compared with placebo (risk ratio 0.84, 95% confidence interval 0.66-1.06).

Pooling risk differences gave the same answer (risk difference -0.01, 95% confidence interval -0.02-0.00), and results were consistent across different doses and disease groups.

In a secondary analysis, only six adverse events were marginally more common on naltrexone than placebo, and they were mild.

Limits

• It left out opioid users, so it doesn't speak to safety in people with opioid use disorder.
• Only 26 of the 89 trials recorded serious adverse events by group, so the main analysis used a minority of the trials.
• It covered tablets only, not the naltrexone shot or implants.

What it means if you're considering naltrexone

For someone worried about serious harm from naltrexone tablets, this is reassuring: across many conditions and doses, serious adverse events were no more common than on placebo. It found only a handful of milder side effects more often on naltrexone, and it doesn't apply to people who use opioids, who were excluded.