Cannabidiol as a novel candidate alcohol use disorder pharmacotherapy: a systematic review.

Abstract

INTRODUCTION

There is burgeoning interest in the therapeutic potential of compounds found in plants of the genus Cannabis. The cannabis plant contains more than 500 constituents and over 100 phytocannabinoids that actively interact with the body’s endocannabinoid system (eCB), of which delta-9-tetrahydrocannabinol (THC) and cannabidiol (CBD) are the most commonly studied. The eCB is an important physiological system with broad activity impacting various functions of the human body, including brain plasticity, inflammation, appetite regulation and learning and memory among others (Aizpurua-Olaizola et al., 2017). It is comprised of at least two G-protein coupled receptors, the cannabinoid type 1 (CB₁) and type 2 (CB₂) receptors (Boggs et al., 2017). THC is a partial agonist of CB₁ and CB₂ receptors (Boggs et al., 2017), and studies using CB₁ receptor antagonists attenuate the effects of THC suggesting responsibility for its psychoactive effects (e.g., Englund et al., 2016; Justinova et al. 2008; Klumpers et al., 2013). Unlike THC, CBD is non-psychoactive which may be attributed to its function as a negative allosteric modulator of CB₁ and CB₂ receptors (Laprairie et al., 2015; Mechoulam et al., 2007). Further impact of CBD on the eCB involves blocking anandamide uptake and inhibiting its enzymatic hydrolysis (Pertwee, 2008; Thomas et al, 2007). There are also numerous non-endocannabinoid signaling systems that CBD may interact with, explaining its diverse biologic effects (Boggs et al., 2017). For instance, CBD can modulate 5-HT_1A receptors (Russo et al., 2005), GPR55 (Ryberg et al., 2007), TRPV1 cation channels (Bisogno et al. 2001) and µ- and δ-opioid receptors (Kathmann et al., 2006).

With numerous and diverse pharmacological effects on a wide variety of body systems, CBD has gained attention for equally numerous health applications. It has demonstrated anti-inflammatory and anxiolytic effects and reports also suggest it may have antipsychotic, antiemetic, antioxidant, and anticonvulsant effects (Hampson et al., 1998; Parker et al., 2002; Zuardi et al., 2006;). However, it is critical to note that this support is derived from a primarily preclinical literature. Despite the ubiquitous health claims of CBD, substantive support only exists in a small number of conditions. In Canada, Sativex (an oromucosal spray of THC and CBD) has been approved as an adjunctive treatment for neuropathic pain in multiple sclerosis (Health Canada, 2007). Similarly, randomized controlled trials of CBD only provide support for seizure disorders (Cunha et al., 1980; Devinsky et al., 2017; 2018a; Theile et al., 2018). Reflecting these findings, earlier this year CBD (Epidiolex ©) was approved by the Food and Drug Administration (FDA) for Lennox-Gastaut and Dravet syndrome, two rare and severe forms of pediatric epilepsy.

Given the salutary effects of CBD in preclinical studies, a variety of clinical applications have been considered, including alcohol use disorder (AUD) and other substance use disorders (SUDs). AUD is characterized by a problematic pattern of alcohol use, which leads to significant impairment and distress (APA, 2013). It is a highly disabling condition and represents a serious public health concern. The FDA has approved three medications for treating AUD including naltrexone (both oral and intramuscular formulation), acamprosate and disulfiram (Leggio & Lee, 2017); however, meta-analytic evidence only supports the use of acamprosate and naltrexone (Jonas et al., 2014). The European Medicines Agency has also approved nalmefene for AUD (Paille & Martini, 2014). These medications have distinct mechanisms of action and are effective for some patients but predicting treatment response is difficult (Jonas et al., 2014). Despite existing pharmacological or psychological therapies, substantial proportions of patients do not have successful outcomes (Aronson, 2015; e.g., Naltrexone number needed to treat (NNT) = 9) and there is a clear need for additional treatments, particularly those targeting alternative novel mechanisms (Litten et al., 2012, 2016a, 2016b), such as the eCB.

The rationale for CBD as an AUD pharmacotherapy comes from both oblique and direct evidence. First, there is evidence that AUD leads to dysfunction in a number of the biological systems for which CBD has favorable effects. For example, alcohol is a potent modulator of the immune system, potentiating alcohol-induced liver inflammation and stimulating immune cells, like monocytes, macrophages, and T lymphocytes, which in turn cause the release of pro-inflammatory cytokines (reviewed in Neupane, 2016). To address these effects, CBD’s anti-inflammatory effects may prove beneficial. Oxidative stress also plays a demonstrable role in potentiating alcohol-related harms (Hernandez et al., 2016; Li et al., 2015) which may be alleviated by CBD’s antioxidant properties. Further, based on the risk for seizures during alcohol withdrawal and persistent disturbances in glutamate neurotransmission (Jesse et al., 2017), CBD’s anticonvulsant effects may have therapeutic potential. Second, preclinical evidence suggests that the eCB plays important roles in motivational properties of alcohol as demonstrated by studies examining CB₁ receptor modulation. More specifically, CB₁ receptor antagonism has been shown to suppress rodent alcohol consumption (Arnone et al., 1997; Colombo et al., 1998; Femenia et al., 2010; Wang et al., 2003) and can block the increased alcohol consumption noted with CB₁ receptor agonist administration (Colombo et al., 2002). Given that CBD may decrease CB₁ receptor activity (through negative allosteric modulation), and CB₁ receptor antagonism has been shown to decrease alcohol consumption in animal models, this evidence further supports the notion that CBD may be a promising treatment for AUD.

The preceding lines of research obliquely suggest that CBD may be a viable AUD pharmacotherapy, but a number of studies have also directly examined the potential role of CBD. The goal of the current systematic review was to characterize the findings from those direct investigations of CBD’s effects on alcohol-related dysfunction or AUD. The broad goal was to appraise the extent to which CBD may be a credible AUD pharmacotherapy. Although a number of existing pharmacotherapies exist for AUD, suboptimal treatment outcomes remain common and the development of novel strategies remains a priority (Kranzler et al. 2018; Litten et al., 2012, 2016a, 2016b).

METHODS

RESULTS

Our initial search generated 303 results, of which 12 met inclusion criteria (see PRISMA flow diagram, Figure 1). In the following sections, we review findings of the effects of CBD on alcohol-related harms in human participants (Consroe et al., 1979; Belgrave et al., 1979; Bird et al., 1980), cell culture (Brenneman et al., 2018) and animal models (Hamelink et al., 2005; Liput et al., 2013; Wang et al., 2017; Yang et al., 2014; Filev et al., 2017; Wang et al., 2017; Viudez-Martinez et al., 2018a; 2018b; Gonzalez-Cuevas et al., 2018). Table 1 provides a concise summary of the alcohol-related harm and associated outcomes of the studies included in this review. Below, the results are according to the most commonly studied outcome domains.

DISCUSSION

The purpose of the current systematic review was to consider the viability of CBD as a novel prospective pharmacotherapy for AUD or specific adverse alcohol-related medical consequences. Overall, the literature was limited to a small number of primarily preclinical studies, and there was an absence of studies in human clinical samples. Nonetheless, the results were consistently supportive of the potential therapeutic benefits of CBD for AUD, particularly in the areas of neurodegeneration, hepatotoxicity, cognition and risk of relapse. Arguably of greatest promise was consistent evidence that CBD is neuroprotective against alcohol-induced brain insults in preclinical models, specifically in the hippocampus and entorhinal cortex (Brenneman et al., 2018; Liput et al.,2013; Hamelink et al., 2005). The selective vulnerability of these two brain regions is notable given that it may lead to the significant behavioural sequelae which go on to impact daily functioning and cognitive ability (Hamelink et al., 2005). Aptly, this leads to the suggestion that neuroprotective compounds may prevent the cognitive deficits that contribute to relapse in some (Stevens et al., 2015).

Direct evidence regarding CBD’s effect on cognition in the context of alcohol-related harms was specific to a favourable effect on impulsive behaviour in animal model of alcohol dependence (Gonzalez-Cuevas et al., 2018). This is a critical finding given that diminished impulse control is thought to play a significant role in the relapsing nature of AUD (Stevens et al., 2015) and this specific form of impulsivity (delay discounting) has been robustly associated with AUD and other forms of addictive behaviour (Amlung et al., 2017; MacKillop et al., 2011). Furthermore, precipitous delay discounting has been found to predict formal addiction treatment outcome (e.g., MacKillop & Kahler, 2009; Sheffer et al., 2012) and natural resolution of alcohol problems (Tucker et al., 2002; 2008).

More generally, the literature examining relapse-like behaviours highlighted the potential benefit of CBD in this domain, particularly in animals that underwent a dependence induction protocol (Gonzalez-Cuevas et al., 2018; Viudez-Martinez et al., 2018a, 2018b). Although the mechanism by which CBD elicits these effects is unclear, CBD prevented reinstatement of alcohol seeking in contexts that trigger relapse while also limiting risk factors associated with relapse (i.e. high anxiety and low impulse control). These findings suggest that CBD may attenuate cue-elicited craving or stress-elicited craving in humans, although that is clearly a hypothesis to be tested. Another preclinical study illustrated that combined CBD and naltrexone was more effective at reducing alcohol consumption than CBD or naltrexone alone (Viudez-Martinez et al., 2018b). They illustrated downregulated gene expression in brain regions responsible for reward and habit formation, relevant to AUD, suggesting a shared putative mechanism of action as demonstrated by the additive effect on treatment outcome (Viudez-Martinez et al., 2018a).

Finally, two small preclinical studies suggest that CBD may have therapeutic potential in alcohol-induced steatohepatitis (Wang et al., 2017; Yang et al, 2014). There was clear consensus among findings from both studies, revealing that markers of liver damage consequent to binge-drinking are prevented by CBD (Wang et al., 2017; Yang et al, 2014). This is promising with the suggestion that perhaps CBD treatment may slow the development of more complicated and severe hepatic problems consequent to alcohol overuse.

The prior findings were restricted to preclinical models, and the only studies identified that utilized human participants used healthy volunteers, not clinical samples. This work revealed that CBD does not alter the acute cognitive effects of alcohol (Belgrave et al., 1979; Bird et al., 1980; Consroe et al., 1975), which does not directly speak to CBD as a treatment for AUD given the single-dose non-therapeutic nature of the studies. On the other hand, these findings provide oblique insights into drug interactions if AUD patients drink while taking CBD, suggesting a relatively tolerable side effect profile of CBD. Ultimately, translational studies in humans exploring these candidate mechanisms identified by the existing preclinical literature are critical to evaluate the potential of CBD as an AUD pharmacotherapy.

The potential therapeutic benefit of CBD can be further derived from the substantial evidence supporting its use in seizure disorders. While the relevance of this indication may seem obscure, other anticonvulsants, specifically topiramate, have shown substantial promise in AUD (Johnson et al., 2003; 2007). In addition to seizure prevention during alcohol withdrawal, topiramate has a large effect on abstinence (g=0.468) and heavy drinking (g=0.406) in AUD as demonstrated by meta-analytic evidence (Blodgett et al., 2014). Moreover, a recent meta-analysis suggested that topiramate was superior to nalmefene, baclofen, naltrexone and acamprosate on alcohol consumption outcomes (Palpacuer et al., 2018). Despite the moderate effect size of topiramate therapy in AUD, side effects (i.e. cognitive impairment, paresthesia, taste abnormalities) reduce topiramates tolerability clinically. This has led to evaluations of other anticonvulsants in AUD. For instance, zonisamide, another anticonvulsant that shares structural features with topiramate, has revealed promising preclinical (Knapp et al., 2007; Sarid-Segal et al., 2009), open-label (Knapp et al., 2010; Rubio et al., 2010) and randomized controlled trial evidence (Arias et al., 2010). However, cognitive impairment of zonisamide may be comparable to topiramate, excluding mental slowing which is only increased in topiramate (Knapp et al., 2015). Levetiracetam is another anticonvulsant that has been examined in AUD. Although it produces fewer adverse effect on cognition (Gomer et al., 2007), research on its efficacy in AUD is not consistent (Fertig et al., 2012; Richter et al., 2012; Sarid-Segal et al., 2008). Other anticonvulsants that have shown promise in AUD include divalproex (Brady et al., 2002) and gabapentin (Leung et al., 2015).

Taken together, the literature signals that CBD may be efficacious in the context of alcohol-related harms by way of its neuroprotective, anti-relapse, and anticonvulsant mechanisms. Despite these potential therapeutic effects, this review highlights the need for further translational studies given that the current literature is limited to a small number of preclinical studies using animal models of AUD phenotypes. It has been previously suggested that CBD may influence specific phases of addiction for different substances of abuse (Prud’homme et al., 2014). Considering the AUD literature, there is merit to exploring CBD’s effects on relapse (Gonzalez-Cuevas et al., 2018; Viudez-Martinez et al., 2018a; 2018b). Moreover, it may prove particularly beneficial in the early stages of treatment given its neuroprotective ability in hippocampal and entorhinal cortical cells (Brenneman et al., 2018; Hamelink et al., 2005; Liput et al., 2013), potential benefit to cognitive function (Gonzalez-Cuevas et al., 2018) and prevention of alcohol-related liver problems (Yang et al., 2014; Wang et al., 2017). Although the animal studies included in this review used robust paradigms for investigating alcohol-related harms, the promise of this literature is from a purely preclinical standpoint. In order to appropriately appraise the utility of a pharmacotherapy, human laboratory and clinical studies (and integrated bench-to-bedside studies) in humans are imperative. More specifically, there is a need for studies that evaluate whether these mechanisms identified in preclinical work will translate to clinical populations. Figure 2 concisely articulates the candidate mechanisms by which CBD may be useful in the treatment of AUD that warrant direct empirical investigation. If positive short-term effects of CBD are present in one or more of these domains, there would be a strong basis for randomized controlled trials in individuals with AUD. However, rather than prioritizing one specific mechanism or phase of treatment, the existing literature suggests investigations that broadly screen the effects in these domains are warranted.

Dosing and route of administration present some of the primary challenges in designing clinical trials to evaluate the therapeutic benefit of CBD. For instance, a majority of the animal studies reviewed above administered CBD transdermally or by injection. This may be relevant as the bioavailability of oral CBD in humans is relatively low, approximately 6% (Agurell et al., 1981). However, recent trials of CBD in the form of Epidiolex© (an oral solution) to treat pediatric seizure disorders may inform dosing and route of administration. These studies revealed that 14 weeks of 10 mg/kg/day or 20 mg/kg/day of Epidiolex© was efficacious in reducing seizure frequency (Devinsky et al, 2017; 2018a; Theile et al., 2018). As a recently approved indication, the FDA recommends a titration schedule involving a starting dose of 5mg/kg/d (administered as doses of 2.5 mg/kg b.i.d); and after one week the dosage may be increased to 10mg/kg/d (5mg/kg b.i.d). In those tolerating this dose but requiring further seizure reduction, a maximum recommended maintenance dose of 10mg/kg b.i.d is recommended (20mg/kg/d)(FDA, 2018). When considering a target dose for AUD, 10 mg/kg/d may be more appropriate for a number of reasons. First, only one of the Epidiolex© seizure trials examined efficacy of both 10mg/kg/d and 20 mg/kg/d and the higher dose provided limited additional benefit in reducing seizure severity (median reduction −42% vs. −37%) (Devinsky et al., 2018a), although these groups were not compared statistically. Moreover, while CBD presents a good safety profile (Bergamaschi et al., 2011), the FDA has recommended a dose adjustment in those with hepatic impairment based on transaminase elevations observed in patients receiving 20mg/kg/d (FDA, 2018). Liver damage is commonly seen in AUD patients (Leggio & Lee, 2017), and the FDA recommends a maximum dosage of 10mg/kg/d for patients with moderate hepatic impairment making this an appropriate target for AUD. Interestingly, the studies on hepatotoxicity as discussed in the current review revealed that CBD was able to prevent development of signs of liver pathology noted with excessive alcohol exposure (Wang et al., 2017; Yang et al., 2014) and safety data from Devinsky et al (2018b) also revealed that patients with elevated transaminase levels received concomitant valproate. An alternative candidate route of administration is inhalation of combusted or vaporized plant material of high CBD composition. However, none of the studies identified in this review used inhalation as a route of administration; all used oral administration, which was tolerated well by participants. Furthermore, variation in topography (e.g., number of puffs, puff duration, inhalation volume and duration) has the potential to create substantial differences in dosing and heterogeneity of effects. Collectively, oral route of administration appears to be most appropriate at this stage.

In addition to the parameters outlined above, the literature discussed in this review identifies target outcomes that would be of interest, including the effects of CBD on alcohol consumption, liver pathology, and cognition. First, measuring abstinence or change in heavy drinking could be quantified by self-report (e.g., timeline followback interview) and could be complemented with more objective biomarkers of alcohol consumption or heavy drinking (e.g., ethyl glucuronide or carbohydrate-deficient transferrin) (Jastrzebska et al., 2016). While serum liver enzymes have low specificity for detecting alcohol use, they would be reliable markers of liver damage. Studies on hepatotoxicity revealed that CBD treatment prevented increases in ALT and AST that were noted in response to heavy alcohol consumption in animals (Wang yet al., 2017; Yang et al., 2014). As such, measuring baseline to endpoint change in these liver enzymes would not only describe the effects of CBD on liver function in AUD patients, but also provide safety data regarding the hepatic impairment noted in Epidiolex© trials. Finally, given the neuroprotective effects of CBD on the hippocampus and entorhinal cortex (Brenneman et al., 2018; Hamelink et al., 2005; Liput et al., 2013), and benefits in impulsive choice (Gonzalez-Cuevas et al., 2018) general changes in neurocognitive function throughout treatment would be of interest.

In sum, this systematic review suggests that CBD may be a credible therapeutic for a wide variety of alcohol-related harms based on the preclinical literature. Several candidate mechanisms by which CBD may produce therapeutic effects in AUD were identified, providing clear avenues for future research. Fundamentally, empirical studies are needed to determine whether these effects translate into favorable outcomes in human preclinical and clinical models and, ultimately, to inform the appropriateness of CBD as a potential pharmacotherapy for AUD.

Acknowledgments

Footnotes

REFERENCES