Brain Correlates of the Alcohol Use Disorder Pharmacotherapy Response: A Systematic Review of Neuroimaging Studies

Abstract

Background: Although Alcohol Use Disorder (AUD) is highly prevalent worldwide, treating this condition remains challenging. Further, potential treatments for AUD do not fully address alcohol-induced neuroadaptive changes. Understanding the effects of pharmacotherapies for AUD on the human brain may lead to tailored, more effective treatments, and improved individual clinical outcomes.

Objectives: We systematically reviewed the literature for studies investigating pharmacotherapies for AUD that included neuroimaging-based treatment outcomes. We searched the PubMed, Scielo, and PsycINFO databases up to January 2021.

Study eligibility criteria, participants, and interventions: Eligible studies included those investigating pharmacotherapies for AUD and employing functional magnetic resonance imaging (fMRI), positron emission tomography (PET), single-photon emission computed tomography (SPECT), and/or proton magnetic resonance spectroscopy (H-MRS).

Study appraisal and synthesis methods: Two independent reviewers screened studies' titles and abstracts for inclusion. Data extraction forms were shared among all the authors to standardize data collection. We gathered information on the following variables: sample size; mean age; sociodemographic and clinical characteristics; alcohol use status; study design and methodology; main neuroimaging findings and brain-regions of interest (i.e., brain areas activated by alcohol use and possible pharmacological interactions); and limitations of each study.

Results: Out of 177 studies selected, 20 studies provided relevant data for the research topic. Findings indicate that: (1) Acamprosate and gabapentin may selectively modulate limbic regions and the anterior cingulate cortex; (2) Naltrexone and disulfiram effects may involve prefrontal, premotor, and cerebellar regions; (3) Pharmacotherapies acting on glutamate and GABA neurotransmission involve primarily areas underpinning reward and negative affective states, and; (4) Pharmacotherapies acting on opioid and dopamine systems may affect areas responsible for the cognitive and motor factors of AUD.

Limitations: Most of the studies were focused on naltrexone. A small number of studies investigated the action of disulfiram and gabapentin, and no neuroimaging studies investigated topiramate. In addition, the time between medication and neuroimaging scans varied widely across studies.

Conclusions: We identified key-brain regions modulated by treatments available for AUD. Some of the regions modulated by naltrexone are not specific to the brain reward system, such as the parahippocampal gyrus (temporal lobe), parietal and occipital lobes. Other treatments also modulate not specific regions of the reward system, but play a role in the addictive behaviors, including the insula and dorsolateral prefrontal cortex. The role of these brain regions in mediating the AUD pharmacotherapy response warrants investigation in future research studies.

Keywords: acamprosate; alcohol; disulfiram; gabapentin; naltrexone; neuroimaging; pharmacotherapy.

Figure 1

PRISMA illustrating the screening process applied in the study.

Figure 2

Neural areas affected by pharmacological treatment compared to placebo in neuroimaging studies. SMA = Supplementary Motor Area—NTX = Naltrexone—ACA = Acamprosate—DSF = Disulfiram—GBP = Gabapentin—FMZ = Flumazenil—ODT = Ondansetron; Mann et al. [27] and Langosch et al. [37]: No areas modulated by ACA; No Deactivation was correlated with NTX [13,14,26,27,28,29,30,31,32,34], ACA [16,38], DSF [39]. Activation was correlated with NTX [33,36] and GBP [40] use. Higher [41] and lower levels of glutamate were found with GBP use [18].

Figure 2

Neural areas affected by pharmacological treatment compared to placebo in neuroimaging studies. SMA = Supplementary Motor Area—NTX = Naltrexone—ACA = Acamprosate—DSF = Disulfiram—GBP = Gabapentin—FMZ = Flumazenil—ODT = Ondansetron; Mann et al. [27] and Langosch et al. [37]: No areas modulated by ACA; No Deactivation was correlated with NTX [13,14,26,27,28,29,30,31,32,34], ACA [16,38], DSF [39]. Activation was correlated with NTX [33,36] and GBP [40] use. Higher [41] and lower levels of glutamate were found with GBP use [18].