The aim of this review was to highlight, based on preclinical literature, the promising therapeutic applications of CBD in the reduction of drinking in AUD, and for improving or preventing alcohol-related damage on the liver and the brain. The main findings on these different topics are displayed in Figure 1 . First, CBD was able to reduce motivation for alcohol, relapse, and the global level of alcohol intake in mice. Next, CBD reduced alcohol-induced liver damage, by reducing liver fibrosis via its immunomodulatory and antioxidant properties, as well as its action on activated HSC, stimulation of autophagy, and via regulation of lipid accumulation in the liver. Last, CBD acts as a multimodal neuroprotective agent that could decrease alcohol-induced neuronal damage leading to cognitive and motor impairment in animals. This latter effect could be associated with CBD antioxidant properties and immunomodulatory action, possibly correlated with the cerebral adenosine system.

Although fewer studies are available to assess the effects of CBD on cannabinoid type 2 receptors (CB2), there could be another mechanism involved in its protective effects on the liver and the brain. CB2 receptors are cannabinoid receptors that are mainly expressed in the immune system (Lotersztajn et al., 2008). CBD seems to have complex interactions with CB2 receptors, acting as a negative allosteric modulator (Martínez-Pinilla et al., 2017).

In an experimental study with cultured hepatic myofibroblasts and activated HSC from human liver biopsy (Julien et al., 2005), CB2 receptors were not detected in normal human liver whereas they were highly up-regulated in cirrhotic liver. Activation of CB2 receptor led to antifibrogenic effects by growth inhibition that probably involved cyclooxygenase-2 (COX-2), and to the increase in apoptosis by regulating oxidative stress. In the same study, mice invalidated for CB2 receptor developed enhanced liver fibrosis.

Hepatoprotective properties of CB2 receptors were also shown in a mice model of carbon tetrachloride-induced acute hepatitis (Teixeira-Clerc et al., 2010). Activation of CB2 receptors reduced liver injury and accelerated liver regeneration by immunomodulation involving TNF-α, IL-6, matrix metallo-proteinase-2 (MMP-2), and reduction of oxidative stress.

In another animal model (Louvet et al., 2011) of alcohol-fed mice, CB2 receptors regulated Kupffer cells polarization by provoking a switch from a classical proinflammatory program of activation (M1) to an alternative anti-inflammatory one (M2). This eventually protected the liver from the deleterious effects of alcohol. Moreover, in the same study, CB2 receptors were shown to reduce steatosis based on paracrine effects of Kupffer cells on hepatocytes.

Finally, as far as the brain is concerned, specific pharmacological activation of CB2 receptors in a forced alcohol consumption rat model rescued alcohol-induced impaired neural progenitor cells (NPC) proliferation, thus counteracting alcohol-induced neuronal damage (Rivera et al., 2015).

However, all these promising findings come from animal models only, and there are currently no results from clinical trials studying CBD in human AUD. It should be noted, however, that one double-blind randomized clinical trial is currently being conducted in the United States. In this ongoing study, CBD is administered versus placebo to patients with AUD, with the aim of reducing the overall level of alcohol drinking (NCT03252756).

Having a similar effect as drugs such as nalmefene (Mann et al., 2016), baclofen (Agabio et al., 2018), or topiramate (Palpacuer et al., 2018), CBD might thus be another good candidate molecule for reducing drinking in subjects with AUD. Furthermore, the antioxidant and immunomodulatory properties of CBD constitute additional and valuable features in the achievement of harm reduction in subjects with AUD, via a reduction or even a prevention of alcohol-related liver or brain damage. While specific pharmacological strategies of harm reduction have previously been developed in other substance use disorders, in particular in opioid use disorder, no other drug has been used in AUD for the specific purpose of reducing alcohol-related damage, even without drinking reduction.

Moreover, CBD seems to have other interesting harm reduction properties, which have not been assessed in AUD models so far, and thus, could not be investigated in this review. CBD has well-known antiepileptic properties: in 2018, the Food Drug Administration (FDA) granted an approval to CBD (EPIDIOLEX^®, GW Pharmaceuticals), for Dravet and Lennox–Gastaut syndromes. Given that patients with AUD are at increased risk to exhibit alcohol-induced or withdrawal-related seizures, CBD could prevent the occurrence or reduce the severity of seizures in this population. CBD also possesses anxiolytic and analgesic properties (Carrier et al., 2006; Rock et al., 2012; Mori et al., 2017). Since subjects with AUD display anxious symptoms or chronic pain more frequently than subjects without AUD (Schuckit and Hesselbrock, 1994; Witkiewitz and Vowles, 2018), CBD could reduce the overall level of anxiety and pain in subjects with AUD, which could improve overall outcomes such as stress and quality of life. Indeed, 5-HT receptors, which are known to regulate anxiety (Ishikawa and Shiga, 2017; Batista and Moreira, 2019; Wang et al., 2019), are one of CBD’s targets (Russo et al., 2005; Pazos et al., 2013), and were studied in AUD (Underwood et al., 2018). For example, ondansetron, a 5-HT3 receptor antagonist, showed some efficacy in both preclinical and clinical studies on AUD (Johnson et al., 2013; Moore et al., 2014; Soyka and Müller, 2017). More recently, a preclinical study in mice (Viudez-Martínez et al., 2018) showed that the 5-HT1a receptor antagonist WAY100635, blocked the positive effect of a CBD-plus-naltrexone combination on motivation and ethanol intake.

Anxiolytic properties of CBD could also be explained by its potential ability to regulate endocannabinoid levels. FAAH is an enzyme responsible for the degradation of endocannabinoids such as anandamide and 2-arachidonoylglycerol (Watanabe et al., 1998; Bisogno et al., 2002; Leweke et al., 2012; Deutsch, 2016), after they have bound to fatty acid binding protein (FABP). CBD inhibits FAAH and thus, prevents anandamide from being degraded (de Filippis et al., 2008; De Petrocellis et al., 2011; Leweke et al., 2012; Stern et al., 2017). Facilitation of endocannabinoid signaling by repeated administration of CBD led to a decrease of chronic stress in mice (Campos et al., 2013). In a human study with a simulated public speaking test in patients suffering from social phobia, CBD was found to significantly reduce anxiety (Bergamaschi et al., 2011). Contrasting with these results, a preclinical study in rats showed an impaired FAAH function in the alcohol-preferring phenotype compared to the nonpreferring phenotype, causing an overreactive endocannabinoid transmission and a compensatory down-regulation of CB1 signaling (Hansson et al., 2007). However, extrapolating all these results in humans seems quite premature: for example, an experimental study on human cells found that CBD had no action on FAAH but rather targeted several types of FABPs (Elmes et al., 2015).

Finally, in addition to hepatic and brain damage, alcohol induces many other noxious effects on the body, for example by inducing alcohol-related myocarditis, or various types of cancers. Because of its immunomodulatory properties, protective effects of CBD against these other harms should be further investigated in both animals and humans. Overall, CBD safety aspects appear to be good, which is another important criterion for extending human research to patients with AUD. So far, no severe clinical states resulting from CBD intoxication have been reported, neither in animal nor in human use. Similarly, to our knowledge, no pharmacological tolerance, withdrawal syndrome, abuse, or addictive behaviors, have been reported hitherto. This is an important factor to consider before using CBD in AUD or other addictive disorders.

Despite the multiple prospects of CBD in AUD that have been emphasized in this review, many issues and unsolved questions remain. The current literature only pertains to animal models, and the translational aspects of the findings listed in this review are yet to be established. Moreover, CBD effective dose range observed in animals is unlikely to be similar in humans. This point is important because the dose–effect relationship of CBD depends on the type of effect and is not always linear. For example, some effects of CBD seem to have an inverted U-shaped dose–response curve. Regarding anxiety, while a dose superior to 20 mg/kg appears to be ineffective in animals (Guimarhes et al., 1990), a human study with a Simulating Public Speaking Test confirmed this U-shaped dose–response, with an efficacy observed with 300 mg of CBD, but not with 150 mg or 900 mg (Linares et al., 2016). However, other animal studies found an anxiolytic effect with repetitive doses of 30 mg/kg, which may activate different pathways (Campos et al., 2013; Fogaça et al., 2018). In animal models of depression, a dose of 30 mg/kg of CBD was found to be as effective as tricyclic antidepressants whereas a 100 mg/kg one was ineffective (Zanelati et al., 2010). With higher doses, activation of TRPV1 reduced the anxiolytic/antidepressant effect (Campos and Guimarães, 2009). Higher doses of CBD (800 mg/kg, 1,000 mg/day) seem to be needed to obtain antipsychotic effect with reduction of positive psychotic score in clinical studies (Leweke et al., 2012; McGuire et al., 2018). Consistent with its large therapeutic target spectrum, sometimes with opposite effects, the therapeutic dose range of CBD should be defined specifically for the various symptoms that clinicians want to alleviate, in connection with hypothetical receptors or secondary pathways.

In conclusion, experimental data underline that CBD offers multiple therapeutic prospects in patients with AUD. CBD seems to facilitate drinking reduction, making CBD an interesting pharmacological option in AUD treatment. Moreover, CBD might provide idiosyncratic protection to the liver and the brain, which could reduce the development and impact of both ARLD and ARBI. In this perspective, CBD treatment could be proposed to subjects who are unable to reduce or to stop alcohol consumption, in order to prevent or reduce the effects of alcohol on the brain and the liver, thus opening new and original therapeutic options for harm reduction in AUD. CBD could have many more positive effects in subjects with AUD, including antiepileptic, cardioprotective, anxiolytic, or analgesic ones. Human studies are thus crucially needed to explore the many prospects of CBD in AUD and related conditions.