¹University of Gothenburg, Institute of Neuroscience and Physiology, Psychiatry and Neurochemistry, Gothenburg, Sweden, ²Karolinska Institute, Department of Clinical Neuroscience, Centre for Psychiatry Research, Stockholm, Sweden, ³Lund University, Faculty of Medicine, Department of Clinical Sciences Lund, Psychiatry, Lund, Sweden, ⁴Lund University, Department of Clinical Sciences, Unit for Biological and Precision Psychiatry, Lund, Sweden, ⁵Linköping University, Centre for Social and Affective Neuroscience, Linköping, Sweden, ⁶Malmö Addiction Centre, Bokgatan 16, Malmö, Sweden

Background: Alcohol Use Disorder (AUD) afflicts 100 million people, causes 3 million deaths/year and generates enormous costs. Available treatments have limited effects. The dopamine deficiency hypothesis of AUD posits that ethanol intake is driven by negative reinforcement generated by a low dopamine state in the brain.

Aims & Objectives: Based on animal experiments we aimed at performing a clinical trial to explore if co-administration of two drugs with supplementary dopamine enhancing actions, varenicline (a partial nicotinic acetylcholine receptor agonist; 1 mg b.i.d.) and bupropion (a weak dopamine reuptake inhibitor; 150 mg b.i.d.) reduces alcohol intake in moderate/severe AUD.

Method: A placebo-controlled thirteen-week Swedish multi-center RCT (n=388; ClinicalTrials.gov NCT04167306, “The COMB-study”) with parallel-group design (placebo+placebo, varenicline+bupropion, varenicline+placebo, placebo+bupropion) was performed in participants with moderate to severe Alcohol Use Disorder (AUD) and with phosphatidylethanol in blood (B-PEth)-confimed ongoing high drinking risk levels. Two primary outcome variables were employed, B-PEth and % heavy drinking days (%HDD). Secondary outcomes were amount of alcohol(g)consumed/day, % abstinent days and craving.

Results: In the modified Intention to Treat (ITT)-analysis (n=384) varenicline+bupropion significantly reduced co-primary outcomes B-PEth (p<0.004, Cohen´s d=0.39) and %HDD (p<0.008, Cohen´s d=0.31), compared to placebo. Varenicline+placebo also reduced B-PEth (p<0.005, Cohen´s d=0.30) and %HDD (p<0.023, Cohen´s d=0.36). In Per Protocol (PP)-analyses effect sizes compared to placebo were larger for varenicline+bupropion (Cohen’s d=0.43 (B-PEth); 0.41 (%HDD)) than for varenicline+placebo (Cohen’s d=0.29 (B-PEth); 0.34 (%HDD)). Secondary outcome alcohol intake (g/day) was reduced in the varenicline+bupropion group only (ITT p<0.002, Cohen´s d=0.44; PP p<0.005, Cohen’s d=0.47), when controlling for multiple comparisons. Varenicline-induced nausea was significantly counteracted by bupropion (incidence, p<0.048; number of days; p<0.010). Across outcome variables (not controlled for multiple comparisons) the two drugs showed complimentary effects in some instances - bupropion, but not varenicline, significantly increased % abstinent days, and varenicline, but not bupropion, significantly reducing B-PEth and craving. Exploratory analyses showed that the combination reduced alcohol intake both in participants with WHO high and very high drinking risk levels. Further, in the PP-population 41% of participants receiving the combination reduced their alcohol intake (g/day) two or more WHO drinking risk levels compared to approx. 23 % in the placebo+placebo group and 32% in the varenicline+placebo and placebo+bupropion groups, respectively, suggesting additive effects. Sex or self-reported heredity did not seem to influence the effect of the combination therapy, whereas the treatment may be most effective in nicotine users (uncertainties due to lack of power).

Discussion & Conclusions: Two brain dopamine elevating treatments (varenicline+bupropion; varenicline+placebo) reduced alcohol consumption compared with placebo+placebo. Effect sizes were largest when varenicline and bupropion were combined and compliance was high (PP-population). Bupropion reduced varenicline-induced nausea. Varenicline+bupropion reduced alcohol intake both in participants with high and very high driniking risk levels and 41% of the PP-population moved two or more WHO drinking risk levels in the combination group compared to 23% in the placebo+placebo group (PP-population).