Abstract
Naltrexone reduces drinking among individuals with alcohol use disorders (AUDs), but it is not effective for everyone. Variability in its effects on reward-related brain activation, genetic variation, and/or cigarette smoking may account for this mixed response profile. This randomized clinical trial tested the effects of naltrexone on drinking and alcohol cue-elicited brain activation, evaluated whether OPRM1 A118G genotype or smoking moderated these effects, and explored whether the effects of medication on cue-elicited activation predicted subsequent drinking. One hundred and fifty-two treatment-seeking individuals with alcohol dependence, half preselected to carry at least one A118G G (Asp) allele, were randomized to naltrexone (50 mg) or placebo for 16 weeks and administered an fMRI alcohol cue reactivity task at baseline and after 2 weeks of treatment. Naltrexone, relative to placebo, significantly reduced alcohol cue-elicited activation of the right ventral striatum (VS) between baseline and week 2 and reduced heavy drinking over 16 weeks. OPRM1 genotype did not significantly moderate these effects, but G-allele carriers who received naltrexone had an accelerated return to heavy drinking after medication was stopped. Smoking moderated the effects of medication on drinking, such that naltrexone was superior to placebo only among smokers. The degree of reduction in right VS activation between scans interacted with medication in predicting subsequent drinking, such that individuals with greater reduction in activation who received naltrexone, but not placebo, experienced the least heavy drinking during the following 14 weeks. These data replicate previous findings that naltrexone reduces heavy drinking and reward-related brain activation among treatment-seeking individuals with AUDs, and indicate that smoking and the magnitude of reduction in cue-elicited brain activation may predict treatment response.