Novel Agents for the Pharmacological Treatment of Alcohol Use Disorder

Disulfiram, the first medication FDA-approved for AUD treatment (in 1951), is an aldehyde dehydrogenase inhibitor that acts by blocking the metabolism of alcohol, increasing acetaldehyde concentration. Acetaldehyde, a toxic metabolite of ethanol, produces an alcohol-induced aversive response, characterized by nausea, vomiting, sweating, flushing, and heart palpitations [23]. Unsurprisingly, these unpleasant effects give disulfiram a relatively poor adherence rate [24, 25]. Disulfiram may also act on dopamine systems, as its major metabolite inhibits the enzyme dopamine beta-hydroxylase (DBH), which aids in metabolizing dopamine into noradrenaline [22, 26, 27]. Serum DBH activity is associated with withdrawal symptoms, and disulfiram has been shown to reduce serum DBH levels [28].

A meta-analysis of 22 RCTs found that disulfiram saw increased success rates compared to placebo in open-label studies only—blinded trials yielded no statistical significance between disulfiram and placebo [29]. However, while these findings do not appear to support the use of disulfiram for treating AUD, this outcome may be due to placebo effects. Research showed that placebo-treated individuals showed decreases in cue-reactivity to alcohol stimuli in a sample of 38 participants [30], and experiencing an acetaldehyde reaction did not necessarily improve treatment response in a sample of 46 participants [31]. Instead, a patient simply being aware of a potential adverse reaction appears to be enough to influence drinking behavior [32].

Disulfiram ingestion under supervision (to ensure adherence) saw significantly better success rates compared to non-supervised treatment, suggesting that supervised administration of disulfiram may still have a place in treating individuals struggling to attain sobriety [33], and unsupervised disulfiram may also be helpful for individuals highly motivated for abstinence. However, adherence management issues limit the utility of disulfiram in the treatment of AUD, and the disulfiram-ethanol interaction can sometimes present as a medical emergency. Therefore, disulfiram is only recommended in the maintenance of abstinence; using this medication to reduce drinking is not advised [34].

While the specific mechanisms through which acamprosate works to treat AUD remain under investigation, it is thought to act on the glutamatergic system as an N-methyl-d-aspartic acid (NMDA) receptor partial co-agonist [35]. This may reduce neuronal hyperexcitability, a phenomenon that occurs in acute withdrawal and protracted abstinence from alcohol.

A meta-analysis of 27 RCTs with a total of 7519 participants found that acamprosate treatment reduced risk of abstinent patients returning to any drinking, but did not reduce rates of binge drinking [36]. A number of trials have also found that acamprosate did not show a significant benefit over placebo [37‐41]. In particular, a large scale trial (COMBINE), which compared acamprosate, naltrexone, and behavioral therapies, both individually and combined with each other, against placebo (N = 1383), found that acamprosate had no significant effect on drinking in comparison to placebo, either alone or in combination with naltrexone and/or behavioral intervention [42]. Placebo effects in this trial may explain some of these negative outcomes, as might differences in trial design and patient characteristics: COMBINE required 4 days of pre-trial abstinence while European trials with positive outcomes were typically conducted in inpatient populations requiring complete detoxification [43]. Overall, however, a Cochrane meta-analysis of 24 RCTs with 6915 participants found that acamprosate significantly reduced the risk of any drinking and increased cumulative duration of abstinence [44].

Acamprosate is generally well tolerated [34] and may also have neuroprotective effects [45, 46]. As chronic alcohol abuse is associated with neuronal changes related to NMDA receptors, this neuroprotection may be particularly important in AUD treatment [47, 48]. Acamprosate is recommended for the achievement and maintenance of complete abstinence, rather than for the reduction of drinking or prevention of relapse in the event of drinking [49]. It is FDA-approved for abstinence maintenance in AUD patients who are abstinent when beginning treatment.

Naltrexone, the best-studied of these three commonly approved medications, was originally approved to treat opioid use disorder. Naltrexone is an antagonist of the mu opioid receptor (with additional affinity for kappa and delta opioid receptors). By attenuating alcohol-induced opioidergic activity in the mesolimbic dopamine system, opioid antagonists like naltrexone, modulate the rewarding effects of alcohol (mediated by dopamine), thereby reducing alcohol consumption [50, 51].

In 1994, the FDA approved oral naltrexone to treat alcohol dependence after two independent 12-week trials, which included 97 and 70 participants, respectively, found that naltrexone significantly decreased drinking days and relapse rates [52, 53]. Additional recent trials demonstrate that naltrexone reduces the rewarding effects of alcohol [54, 55], alcohol craving [40, 56], drinks per drinking day [57], and relapse rates [58, 59], strengthening the initial findings.

Extended-release injectable naltrexone (i.e., Vivitrol), administered once monthly by a medical professional, may be beneficial for individuals who are more sensitive to naltrexone’s adverse side effects or have difficulty adhering to oral medication [60]. A six-month multisite trial of 380 mg injectable naltrexone in 624 patients with alcohol dependence found significant reductions in heavy-drinking days versus placebo [61]. However, evidence for naltrexone remains mixed. Another RCT reported no significant differences between naltrexone and placebo (N = 183), and that the clinical utility of naltrexone was limited by its adverse effects [62]. Additionally, clinical trials of naltrexone often yield modest effect sizes [63]. A meta-analysis of 53 naltrexone RCTs with a total of 9140 participants found naltrexone to significantly reduce both the risk of return to any drinking and return to binge drinking [36]; however, both of these associations were modest in magnitude [64].

Oral and injectable naltrexone show similar decreases in likelihood of binge drinking [65] and both are generally well tolerated, with fairly mild side effects [66]. Importantly, however, naltrexone does block the therapeutic effects of opioid analgesics and can precipitate opioid withdrawal in patients who have developed physical dependence to opioids [34]; therefore, individuals who are prescribed naltrexone for AUD must be monitored for opioid use and withdrawal. Of note, naltrexone is contraindicated for patients with acute hepatitis or liver failure, and should be “carefully considered” in patients with acute liver disease [67], potentially limiting its use in the alcohol-associated liver disease (AALD) population. In summary, naltrexone appears to have a moderate effect on the reduction of alcohol use.

Nalmefene works in a similar manner to naltrexone as an antagonist at the mu and delta opioid receptor, but is also a kappa opioid receptor partial agonist [68]. Via its kappa agonist activity, particularly centered in dopaminergic nucleus accumbens circuitry, nalmefene may reduce motivation for self-administration and withdrawal-induced alcohol consumption [69, 70].

Nalmefene was approved by the EMA in 2013 for the reduction of alcohol consumption among patients with AUD. Approval followed findings from three multicenter 6-month clinical trials enrolling 604, 667, and 718 individuals, respectively, in which participants took the medication or placebo on an as-needed basis [71‐73]. In these trials, nalmefene decreased total alcohol consumption and heavy-drinking days. However, the drug’s approval based on these studies has received criticism due to limited evidence of efficacy, especially as these trials were conducted only against placebo rather than an active medication comparison (e.g., naltrexone) [34, 74]. A more recent meta-analysis of the efficacy of nalmefene, which included five RCTs (N = 2567), found that participants treated with nalmefene had 1.65 fewer heavy-drinking days per month than participants treated with placebo after 6 months [75].

Studies indicate that nalmefene is associated with more adverse events and study dropouts compared to placebo, and the most frequently reported of these are dizziness, nausea, vomiting, insomnia, and headache [36, 76]. Overall, while nalmefene may reduce heavy-drinking rates, its effects on other outcomes [77] remain small-to-moderate, and study withdrawals related to adverse events are common in nalmefene trials. However, similar to naltrexone, this medication may appeal to patients with goals of reducing alcohol consumption and those reluctant to engage in abstinence-based treatment.

Baclofen acts on the γ-amino butyric acid (GABA) system as a GABA_B agonist. It was approved for treatment of AUD in France in 2018 [78] and has been used off-label for AUD for over a decade in other countries, especially other European countries and Australia [79, 80].

Clinical trials of baclofen have produced mixed results [65‐67]. A recent meta-analysis of 12 RCTs (N = 703) showed that baclofen in comparison to placebo was associated with higher abstinence rates [81‐83]; however, it did not increase abstinent days or decrease craving, heavy drinking, depression, or anxiety [84]. Another meta-analysis of 13 RCTs (N = 1492) indicated that baclofen (again vs placebo) was associated with longer time to relapse and a larger percentage of abstinent patients [85]. Furthermore, greater alcohol use at baseline was correlated with a greater treatment effect [85]. In contrast, however, another recent meta-analysis of 12 RCTs with 1128 total participants found no significant differences between baclofen and placebo in primary (i.e., percent abstinent days, percent heavy-drinking days, return to any drinking, and study dropout) or secondary outcomes of interest (i.e., anxiety and craving); however, baclofen did increase depression and adverse effects including sedation and vertigo [86].

Baclofen’s significant adverse side effects include drowsiness, sedation, headache, vertigo, confusion, perspiration, muscle stiffness and/or abnormal movements, slurred speech, and numbness [86‐88]. Tolerance to baclofen may also develop with chronic administration [89, 90]. Additionally, dose and sex may moderate baclofen tolerability and response; escalation of dosage in response to developing tolerance can increase sedative side effects, which affect women significantly more than men at the same dose [91]. Importantly, cessation or reduction in dose can precipitate potentially life-threatening withdrawal syndrome [90].

The variability in baclofen’s effectiveness seen across studies may be partially explained by high baclofen pharmacokinetic variability seen among individuals with AUD. This heterogeneity is an important factor to take into account when considering baclofen as an AUD treatment [92]. Of note, there is also some evidence that baclofen might be particularly useful in treatment of AUD among individuals with liver disease [93, 94].

In summary, baclofen seems to effectively promote abstinence; however, it shows mixed results regarding its clinical effects on non-abstinence outcomes (e.g., reduction in heavy drinking), significant adverse side effects, and inter-individual variability in response. As such, baclofen as a treatment of AUD—as well as its optimal dosage—continues to be debated [95].

Sodium oxybate (SMO), the sodium salt of gamma-hydroxybutyrate (GHB), has been utilized as a medication for a number of disorders. SMO is approved in Italy and Austria for the treatment of AUD [96]. SMO acts on the GABA system both directly as a GABA_B partial agonist and indirectly through GHB-derived GABA [97, 98].

A Cochrane meta-analysis of 13 RCTs with a total of 649 participants found that SMO (50 mg) was effective compared to placebo in the treatment of alcohol withdrawal syndrome (AWS) and in preventing relapses in previously detoxified participants, and that SMO was more effective than naltrexone and disulfiram in maintaining abstinence [99]. Another recent meta-analysis found that SMO, compared to placebo, increased abstinence rates by up to 34% in a sample of 711 participants with very high drinking risk levels [100]. However, use of GHB as a recreational drug raises concern regarding the abuse potential of SMO. Indeed, in another clinical study with a sample of 48 participants, craving for SMO was observed in 10% of participants with psychiatric comorbidities [101], indicating that patients may be at risk of developing craving for and abusing SMO.

SMO may be particularly effective in combination with other pharmacotherapies. It was shown to reduce relapses when administered with naltrexone and escitalopram [102]. A study of 52 subjects with chronic, treatment-resistant alcohol dependence found that participants stayed in treatment significantly longer when SMO was co-administered with disulfiram than with SMO alone [103]. Additionally, in a case study of seven partial- and non-responders to SMO treatment, co-administration with nalmefene was effective in promoting abstinence, reducing heavy-drinking episodes, and importantly, suppressing craving for SMO [104].

SMO craving and abuse potential may limit its use; however, at therapeutic doses, SMO abuse seems to be a relatively infrequent phenomenon among patients without psychiatric comorbidities or poly-drug use [96, 100, 105]. Safety data show that serious adverse events with SMO treatment are rare, with the most common adverse effects being transitory dizziness and vertigo [96]. Overall, the effectiveness of SMO lies largely in reducing withdrawal symptoms and increasing abstinence rates. Its therapeutic use remains controversial due to its potential for craving and abuse, and further study on co-administration with other pharmacotherapies appears promising.

Despite that the anti-convulsant medication topiramate has not been approved by the FDA or EMA for AUD treatment, the US Department of Veterans Affairs recommends topiramate as a treatment for AUD, and it has been suggested by the American Psychiatric Association (APA) for the pharmacological treatment of patients with AUD [106, 107]. Currently, the specific mechanisms of action of topiramate remain under investigation. However, the drug is thought to inhibit glutamate α-amino 3-hydroxy-5-methyl-4-isoxazoleproprionic acid (AMPA) and kainate receptors [108‐110] and L-type calcium channels [111], as well as enhance the inhibitory activity of GABA [109]. These effects, taken together, work to attenuate dopaminergic activity mesolimbic reward circuits, thereby reducing both alcohol craving and withdrawal symptoms [112].

Preclinically, topiramate was seen to decrease ethanol consumption in rodent models [113‐115]. Clinical studies have found that topiramate reduced craving and alcohol drinking in a sample of 150 participants over 12 weeks [112], and decreased drinking days, heavy-drinking days, and number of drinks per day in a sample of 85 participants in a 14-week trial [116]. A large, multisite 14-week RCT that enrolled 371 participants with alcohol dependence found reductions in heavy-drinking days and improvements in various self-reported drinking-related outcomes [117]. A meta-analysis including seven RCTs of topiramate for AUD with a total of 1125 participants found that topiramate significantly increased number of days abstinence and decreased heavy-drinking days compared to placebo [118]. Of interest, this meta-analysis also directly compared effect sizes for topiramate and naltrexone, finding that topiramate was significantly more effective than naltrexone in reducing binge drinking, but not in improving abstinence [118]. Another recent month-long study of 94 patients that directly compared topiramate and baclofen demonstrated that baclofen was overall more effective (i.e., 61% of participants receiving baclofen achieved abstinence vs 38% with topiramate) and better tolerated (i.e., 33% of participants receiving topiramate dropped out of the study vs 24% with baclofen) than topiramate [119].

The potential of reduced tolerability and significant side effects such as pruritis, paresthesia, anorexia, dysgeusia, dizziness, nervousness, and cognitive impairment, including difficulty with concentration and attention, are concerns for the use of topiramate in the treatment of AUD [50]. These cognitive impairments were found to be dose-associated and include impairments in working memory and verbal fluency, as well as mental slowing [116]. Topiramate’s side effects, especially those affecting memory and cognition, present reasons for concern and warrant long-term trials to further investigate these effects. Slow dose titration and ongoing patient monitoring is also recommended.

It is worth noting a potential strength of topiramate: the possibility of starting treatment while patients are still actively drinking, allowing topiramate to serve as a harm-reduction or abstinence-initiation treatment, rather than only being used to prevent relapse in already detoxified patients [120]. Topiramate has already been demonstrably effective in clinical applications across a number of drinking outcomes [121], but longer-term trials of topiramate are recommended to optimize dose and duration, as well as further exploring side effects.

Another anti-convulsant medication, gabapentin, is also suggested by the APA for use in AUD treatment with the goal of reducing alcohol consumption or achieving abstinence [107]. It is thought to modulate GABA activity by indirectly interacting with voltage-gated calcium channels [122]. Preclinical studies of gabapentin’s effects on ethanol consumption have shown mixed results, with some studies finding that gabapentin reduced ethanol intake in alcohol-dependent rodents [123] and others finding that it increased self-administration and binge-like drinking [124, 125].

Gabapentin has shown promising results in human laboratory studies and clinical trials. Gabapentin reduced percent heavy-drinking days and drinks per day, as well as increased percentage abstinent days in a 28-day trial with 60 participants [126]. In another 10-day study with 21 participants, gabapentin significantly delayed return to heavy drinking [127]. In a proof-of-concept one-week human laboratory trial (N = 33), participants treated with gabapentin showed significant decreases in alcohol craving in comparison to placebo [128], and a follow-up 12-week RCT with 150 participants found that gabapentin significantly increased abstinence rates and percentage of no heavy-drinking days [129]. Gabapentin may also be particularly effective in combination with other medications. For instance, a combination of gabapentin and flumazenil over 39 days in 60 participants increased percentage of abstinent days and delayed time to first heavy-drinking day compared to placebo. This trial also showed an interaction effect with pre-study alcohol withdrawal, such that those with more severe withdrawal symptoms at baseline benefitted more from the treatment during the trial [130]. In combination with naltrexone, gabapentin was also shown to be more effective (i.e., decreased heavy-drinking days, fewer drinks per drinking day, and delayed time to heavy drinking) compared with naltrexone alone in a 6-week trial with 150 participants [131].

A Cochrane review of 25 studies (N = 2641) found anticonvulsants, including gabapentin, to significantly reduce both heavy-drinking days and drinks per drinking day in comparison to placebo. Anticonvulsants were also associated with longer time-to-relapse and fewer heavy-drinking days compared to naltrexone [132]. A recent meta-analysis of 7 RCTs with a total of 751 participants found that gabapentin compared to placebo only significantly decreased percent heavy-drinking days, although trend-level effect estimates were shown for other alcohol-related outcomes (i.e., relapse, drinks per day, complete abstinence, percent days abstinent, and concentration of gamma-glutamyl transferase (GGT)) [133]. In a 16-week RCT with 96 participants, gabapentin improved the percentage of individuals with no heavy-drinking days and increased total abstinence rates over placebo, especially among participants with more severe pre-study withdrawal [134]. However, another recent 6-month multisite RCT of extended-release gabapentin conducted in 346 participants with at least moderate AUD found no effects over placebo for any clinical outcomes including abstinence rate, percent heavy-drinking days, drinks per drinking day, or alcohol craving [135]. While these findings may be related to potential pharmacokinetic issues relating to the specific formulation of extended-release gabapentin used in the trial, it is also possible that gabapentin may simply be more effective in patients with more clinically relevant and severe symptoms of alcohol withdrawal [130, 134].

The most common adverse events associated with gabapentin treatment are somnolence, dizziness, peripheral edema, and ataxia or gait disorder [136]. Notably, gabapentin does carry the potential for misuse and abuse, particularly in individuals with opioid use disorders [137]; therefore, recommendation of gabapentin to individuals with comorbid substance use disorder should be carefully considered. In summary, gabapentin shows some clinical efficacy, especially in populations with more severe withdrawal symptomology, but more extensive investigation is recommended.

Varenicline is a nicotinic acetylcholine receptor agonist (partial at α4β2; full at α7) that is used for the treatment of nicotine dependence. In rodent models, varenicline was shown to reduce ethanol seeking, intake, and binge-like consumption [138‐140]. Additionally, the combination of varenicline and naltrexone decreased alcohol drinking more effectively than either drug on its own [141].

Clinically, varenicline was found to significantly reduce weekly percent heavy-drinking days, drinks per day, drinks per drinking day, and alcohol craving compared to placebo in a 13-week multisite RCT with 200 participants [142]. However, a 12-week RCT (N = 160) found no effect of varenicline over placebo on heavy-drinking days [143], and a recent 6-week human laboratory study in 47 participants found that varenicline did not significantly attenuate cue-induced alcohol craving relative to placebo [144].

Varenicline appears to be especially relevant to heavy-drinking smokers. Approximately 20%–25% of current smokers are estimated to also be heavy drinkers [145]. A human laboratory study with 20 heavy-drinking smokers on varenicline for 7 days found that varenicline significantly reduced the number of drinks consumed and increased the likelihood of complete abstinence in the human laboratory paradigm [146]. Moderator analyses of data collected in the above multisite trial [142] indicated that varenicline was more efficacious in reducing drinking among smokers who also reduced their cigarette smoking [147] and among individuals with lower severity of AUD, such that that varenicline significantly reduced percent heavy-drinking days and drinks per drinking day in low-severity individuals, while the most severe group showed no differences between varenicline and placebo on drinking outcomes [148].

Findings also indicate that varenicline may be more effective as an AUD treatment in men, as a 16-week RCT with 131 participants found that varenicline combined with medication management decreased percent heavy-drinking days only in men, but improved smoking abstinence in both men and women [149]. Varenicline is well tolerated, suggesting that it may serve as a promising option for AUD treatment, especially in the case of individuals co-using alcohol and nicotine.

Aripiprazole is an antipsychotic drug that acts as a partial agonist at the dopamine D2 and 5-HT1A receptors and as an antagonist at the 5-HT2A receptor [150]. Preclinically, aripiprazole has been shown to reduce ethanol-induced place preference and decrease drinking behaviors in mice [151], as well as reducing alcohol consumption in alcohol-preferring rats [152, 153].

A human laboratory study in 18 healthy participants indicated that aripiprazole affected outcomes related to alcohol consumption (i.e., subjective response to alcohol), including reducing the euphoric effects of alcohol and increasing sedative effects [154]. In a sample of 30 participants with AUD, aripiprazole compared to placebo, was found to attenuate cue-induced neural activation in the ventral striatum, a region associated with reward [155]. Another recent clinical laboratory study in which 99 participants with AUD took aripiprazole over eight days found that aripiprazole reduced the number of drinks consumed in a bar lab setting, especially among individuals with low self-control, and prolonged the latency to drink in individuals with high impulsivity [156].

Results from clinical trials of aripiprazole are mixed. A 16-week RCT directly comparing aripiprazole against naltrexone (N = 75) found that the two medications were overall comparable in reducing heavy-drinking days and increasing days of abstinence, although patients treated with naltrexone had reduced craving compared to aripiprazole [157]. A 35-day open-label trial assessing the combination of aripiprazole and topiramate in 13 heavy-drinking participants found significant reductions in alcohol use. Additionally, this trial provided no evidence that the side effects of the two medications are additive, indicating that the combination is generally safe and well tolerated [158]. However, another recent five-week study assessing topiramate, aripiprazole, and their combination with 90 participants found that effects on drinking reduction were due to topiramate, while no significant findings were seen for aripiprazole for any outcomes [159]. Another multicenter RCT enrolling 295 participants over 12 weeks found no significant difference between the aripiprazole and placebo groups in percentage of participants completely abstinent from alcohol throughout the study or time to first drinking day; however, the average number of drinks per drinking day was significantly lower for aripiprazole [160]. The aripiprazole group yielded significantly higher discontinuation rates and earlier discontinuation, largely due to adverse events associated with side effects, especially when dosage exceeded 15 mg/day. The most common side effects cited in cases of discontinuation were fatigue, insomnia, restlessness, anxiety, and deficits in attention. Of note, long-term use of antipsychotics like aripiprazole may be associated with more severe adverse effects such as tardive dyskinesia, with risk factors including older age and female sex [161, 162].

In summary, these findings suggest that aripiprazole may be more effective at lower doses and in more impulsive individuals, although larger confirmatory studies are needed to pursue these personalized medicine approaches.

Ondansetron is a 5-HT3 antagonist that is used to treat nausea and vomiting. Although the specific mechanism of action remains under investigation, ondansetron may address serotonergic dysfunction common in early-onset AUD [163, 164], and may reduce alcohol craving via 5-HT3 projections to dopaminergic connections in the midbrain. Preclinically, 5-HT3 antagonism has been shown to block acquisition and maintenance of ethanol self-administration [165] and reduce ethanol-associated dopamine concentration in the nucleus accumbens [166]. Ondansetron was also found to block the development and expression of sensitization to the locomotor stimulant effects of ethanol [167] and reduce voluntary ethanol intake, preference, and withdrawal seizures in rodent models [168, 169].

Clinically, ondansetron may be particularly effective in combination with naltrexone. In an 8-week RCT in 20 participants with early-onset AUD, ondansetron and naltrexone (in comparison with placebo) significantly reduced drinks per drinking day and trended towards an increase in percentage of days abstinent [170]. Another combination study in 90 participants after 7 days on ondansetron and naltrexone found that the combination decreased craving for alcohol and ventral striatal activation to alcohol cues [171].

Ondansetron may also be suitable for individuals with biological predisposition to early-onset AUD. In an 11-week RCT of 271 participants, ondansetron was shown to reduce self-reported drinking such that patients with early-onset AUD who received ondansetron reported fewer drinks per day and more days of abstinence compared to placebo [172]. Ondansetron, combined with cognitive behavioral therapy in an 8-week open-label trial (N = 40) comparing effects in early-onset versus late-onset AUD, found that drinks per drinking day and alcohol-related problems (i.e., Drinker Inventory of Consequences; DrInC) were significantly decreased among those with early-onset AUD compared to those with late-onset AUD [173]. Furthermore, in an 11-week study with 253 participants, ondansetron at 4 μg/kg reduced overall craving significantly in participants with early-onset AUD, while a lower dose (1 μg/kg) reduced craving in participants with late-onset AUD. These reductions in craving were also associated with reduced drinking (i.e., drinks per drinking day) and an increased percentage of abstinent days [174].

Ondansetron is well tolerated with relatively mild side effects including diarrhea, constipation, and headache [175]. Overall, these studies suggest a potential role for ondansetron as an AUD treatment, especially in participants with early-onset AUD and possibly in combination with naltrexone.

Mifepristone (RU-486), a glucocorticoid receptor antagonist, works on the stress system by regulating the amygdala [176]. Preclinically, both systemic and central amygdala injections of mifepristone were shown to suppress yohimbine stress-induced reinstatement of alcohol seeking, indicating that the central amygdala plays an important role in mifepristone’s effects on ethanol-seeking [177, 178]. Recent preclinical research in primates has shown mixed results, such that mifepristone was shown to decrease chronic voluntary alcohol consumption in rhesus macaques at doses of 30 and 56 mg/kg per day [179], but had no effect on alcohol-seeking or self-administration in baboons at slightly lower doses of 10–30 mg/kg per day [180]. Of note, in the former study, cessation of mifepristone treatment resulted in a rapid return to baseline intake levels, and mifepristone was not effective in preventing a relapse during early abstinence.

In a human laboratory study with 56 alcohol-dependent participants, mifepristone (taken for one week) was effective in reducing alcohol craving and consumption relative to placebo, improved liver-function markers, and was overall well tolerated [178]. A two-week Phase 4 RCT examining the effects of mifepristone on cognition in AUD was recently conducted, but recruitment challenges rendered the results inconclusive [181, 182]. Of note, in this trial, participants who received mifepristone had higher Beck Depression Inventory scores compared to placebo at 4 weeks post-randomization despite similar scores at baseline between the two groups, indicating a greater severity of depression symptoms caused by the medication. Additional clinical research on mifepristone as a treatment for AUD and its potential side effects is warranted.

Ibudilast is an inhibitor of phosphodiesterases (PDE)-3, -4, -10, and -11 [183] and macrophage migration inhibitory factor (MIF) [184]. Ibudilast has been shown to dose-dependently suppress pro-inflammatory cytokines, such as interleukins IL-1β, IL-6, and tumor necrosis factor alpha (TNF-α), and to increase the anti-inflammatory cytokine IL-10 and neurotrophic factors [185]. As increases in inflammation are seen in AUD [186], the effects of ibudilast in treating AUD are thought to be driven by its anti-inflammatory and pro-neurotrophic qualities [187].

Preclinically, ibudilast was demonstrated to reduce alcohol intake in two rat models, and decreased drinking selectively in alcohol-dependent mice in comparison to non-dependent mice [188]. These preclinical findings align with prior rodent studies in which pharmacological inhibition of PDE also reduced alcohol intake [189‐191].

In a 7-day human laboratory crossover trial (N = 24), ibudilast was well tolerated and decreased tonic craving in comparison to placebo. Additionally, ibudilast improved mood during exposure to alcohol and stress cues, and reduced the mood-altering and stimulant effects of alcohol among participants with more severe depressive symptoms [192]. Another recent 2-week RCT enrolling 52 participants found that ibudilast also significantly decreased the odds of heavy drinking during the trial by 45% compared to placebo and attenuated neural response to alcohol cues in the ventral striatum [193, 194].

Ibudilast appears to be well tolerated [195], with common adverse side effects including gastrointestinal symptoms (nausea, vomiting, abdominal pain, and diarrhea), headaches, and depression [192, 196]. In the aforementioned 2-week RCT, no significant differences in side effects were seen between groups [193]. In summary, early findings from clinical studies of ibudilast for AUD treatment appear promising and warrant further clinical investigation.

Prazosin and doxazosin are alpha-1 adrenergic receptor antagonists with similar chemical structures that can readily cross the blood-brain barrier and block noradrenergic excitation of the mesolimbic dopaminergic system [197, 198]. Both medications are used for the treatment of hypertension and benign prostatic hyperplasia. While these medications show good safety and tolerability, doxazosin appears to have a better clinical profile, such as improved absorption profile and a longer half-life, leading to less frequent dosing. Adrenergic receptors regulate sympathetic nervous system activity through activation of the neurotransmitter norepinephrine [199]. Stress physiology is disrupted with chronic alcohol use, particularly during early alcohol abstinence. In early abstinence, individuals with AUD experience more emotional dysregulation, stress, and alcohol cravings, all of which can increase the risk of relapse. Thus, alpha-1 blockers like prazosin and doxazosin may help to normalize these stress system changes seen in AUD [200].

Preclinical work found that prazosin reduced ethanol-related operant responding in acute withdrawal for dependent, but not non-dependent rodents [201]. In addition, prazosin treatment prevented yohimbine stress-induced reinstatement of ethanol seeking [202] and attenuated ethanol intake during relapse in alcohol-preferring rats [203]. Doxazosin decreased voluntary alcohol intake in alcohol-preferring rats in a dose-dependent manner, and this effect was likely not due to general motor impairment [204]. Further, doxazosin significantly reduced voluntary ethanol intake in a rodent model of AUD and stress exposure [205].

In humans, an early 6-week pilot study with 24 randomized participants with AUD revealed that prazosin treatment was associated with fewer drinking days per week and fewer drinks per week than placebo [206]. More recently, 92 participants with AUD completed a 12-week double-blind study with prazosin and medication management [207]. Results from intent-to-treat analyses showed that prazosin participants had greater reductions in heavy drinking and rates of drinking over time, although these effects were modest. A 10-week RCT comprising 41 individuals with AUD showed no significant effect of doxazosin over placebo on quantity of alcohol consumption (i.e., drinks per week, heavy-drinking days) [208]. However, further examination of clinical moderators revealed that doxazosin significantly reduced drinks per week and heavy-drinking days among individuals with higher family history of AUD and higher standing diastolic blood pressure [209]. Similarly, moderator analyses from a 12-week RCT of prazosin (N = 100) found that one’s degree of alcohol withdrawal symptoms predicted clinical response, such that participants with high levels of withdrawal symptoms benefited the most from treatment (e.g., reduced craving and heavy-drinking days; improved mood symptoms) [210]. A small meta-analysis of 6 studies with a total of 319 participants tested the effectiveness of drugs acting on adrenergic receptors for AUD and found a significant treatment effect of prazosin and doxazosin on alcohol consumption but not abstinence [211]. In summary, prazosin and doxazosin show some early promise as a treatment for AUD and may be particularly beneficial as a harm-reduction approach and for individuals with significant alcohol withdrawal symptoms or family history of AUD as well as comorbid post-traumatic stress disorder (PTSD).

Alcohol use has been shown to alter the glutamate system. Chronic and binge drinking inhibit glutamate levels and transmission through blockade of NMDA receptors, subsequently leading to elevated glutamate levels during alcohol withdrawal [212‐214]. N-Acetylcysteine (NAC) is a cysteine prodrug that works to restore glutamatergic tone in reward circuitry by improving the expression and function of the cysteine-glutamate exchanger and normalizing glial glutamate transporters [215, 216].

Preclinically, NAC has been shown to reduce withdrawal effects [217], block the development of behavioral sensitization to alcohol [218], and attenuate biological adaptations induced by alcohol cessation (i.e., blocked the reduction of transcription factor ΔFosB in the nucleus accumbens; reduced corticosterone and leptin levels) [218‐220]. Additionally, NAC reduced ethanol-seeking and self-administration in rodents [215, 216]. NAC and aspirin co-administration also reduced ethanol intake and relapse binge drinking in ethanol-preferring rats [221].

In a recent meta-analysis of seven RCTs with a total of 245 participants, NAC compared to placebo was shown to reduce craving symptoms across a number of substance use disorders [222]. A secondary analysis of a 12-week, multisite RCT of NAC to treat cannabis use disorder in 277 participants found that NAC increased odds of between-visit abstinence, and reduced alcohol consumption (i.e., drinks per week and drinking days per week) by 30% [223]. However, a recent 5-day human laboratory study (N = 9) found that NAC did not attenuate alcohol self-administration [212]. Additional clinical trials will also examine the effectiveness of NAC in adolescent and adult samples of AUD ([224], NCT03707951). These trials include samples with comorbid psychopathology (i.e., PTSD) and will use neuroimaging methods to examine the neural circuitry underlying NAC’s modulation of relevant metabolites and neural reactivity to alcohol cues.

Orally-administered NAC appears to be well tolerated, with the most common adverse effects being nausea and diarrhea [212, 222]. Of note, NAC is currently being tested in adolescent AUD, a unique prospect as no other AUD pharmacotherapies are yet approved for adolescents [225]. In summary, NAC represents a promising potential treatment for AUD and merits further exploration.

Suvorexant is a dual orexin antagonist that is used for the treatment of insomnia. The orexin/hypocretin system is well known for its role in sleep-wake regulation but has more recently been implicated in AUD [226]. Orexins are neuropeptides that are densely localized in the lateral hypothalamus. Orexins A and B bind to the G-protein coupled orexin-1 and orexin-2 receptors. Orexin A has equal affinity for both receptors, whereas orexin B has selectivity for orexin-2 receptors. The dense orexin projections from the lateral hypothalamus to the ventral tegmental area provide neurobiological support that orexins may influence responses to rewarding stimuli, including alcohol [227, 228].

Animal models demonstrate that orexin-1 receptor antagonists reduce alcohol drinking in alcohol-dependent mice and alcohol-preferring rats [229‐233]. Orexin-2 antagonists also reduce alcohol drinking and reinstatement/relapse in mice and rats [234‐236]. Similarly, dual orexin antagonists reduce alcohol consumption in alcohol-preferring rats [237, 238]. Given the effectiveness of orexin antagonists in reducing alcohol drinking at the preclinical level of analysis, suvorexant has garnered interest as a drug that can be repurposed to treat AUD [239]. While no animal or human laboratory study has directly examined the efficacy of suvorexant on alcohol-related behaviors, two ongoing Phase 2 RCTs will assess suvorexant’s potential as a treatment for AUD (NCT04229095) and comorbid AUD + insomnia (NCT03897062). The sedative effects of suvorexant are of notable concern, particularly if individuals engage in alcohol drinking during treatment [240]. Thus, the level of patient monitoring throughout the treatment phase and finding the optimal time of dosing to negate the additive sedative effects are important factors to consider to fully evaluate its therapeutic potential.

ABT‐436 is an orally active, highly selective vasopressin type 1B (V1B) receptor antagonist. V1B antagonists attenuate basal hypothalamic-pituitary-adrenal (HPA) axis activity and have shown favorable effects in rat models of alcohol dependence, including attenuating reinstatement of alcohol self-administration [241], and diminishing alcohol intake by alcohol-preferring and alcohol-dependent rats [241, 242].

ABT-436 has been shown to attenuate basal HPA axis activity in humans [243]. A 12-week RCT enrolling 150 participants found ABT-436 to be associated with an increased percentage of days of abstinence compared to placebo, as well as significantly reduced cigarette use [244]. However, no differences were found on heavy-drinking days or alcohol craving. A subgroup analysis also showed that ABT-436 appeared to have greater efficacy among participants with high baseline stress levels. This study was the first Phase 2 clinical trial that tested a V1B receptor antagonist for AUD.

The finding that ABT-436 reduced both alcohol drinking and smoking indicate a potential use for V1B antagonists to treat co-use of alcohol and nicotine. ABT-436 is generally well tolerated in humans, with the most common side effect being diarrhea. Furthermore, results indicate that patients with high levels of stress may specifically benefit from medications targeting the vasopressin receptor.

N-[(4-Trifluoromethyl) benzyl] 4-methoxybutyramide (GET73) is a GHB analogue that has shown promising in vitro and in vivo preclinical results as a potential agent for the treatment of AUD. The addition of GET73 to cultures of rat hippocampal neurons rescued negative ethanol-induced effects, reductions in cell viability, and increases in reactive oxygen species production, providing evidence for a neuroprotective role of GET73 as an AUD treatment [245]. In vivo, GET73 treatment at low, non-sedative doses (5–50 mg/kg) reduced alcohol intake and suppressed relapse in alcohol-preferring rats, as well as exerting anxiolytic effects [246]. These effects are similar to those seen with GHB administration [247]; however, GET73 was shown not to bind to either high- or low-affinity GHB binding sites in rat cortical membranes [246]. Recent studies indicate that GET73 may act as a negative allosteric modulator at the metabotropic glutamate subtype 5 receptor (mGluR5) [248]; however, the complete mechanism of action of GET73 remains unclear.

A translational study examining GET73-alcohol interactions found that neither 30 nor 100 mg/kg GET73 administered in rats (equivalent to doses employed in humans) potentiated alcohol-induced intoxication. Additionally, GET73 administered both in the presence and absence of alcohol was well tolerated in two samples of 14 and 11 participants, with no severe adverse events and no difference in adverse events [249]. A Phase I clinical trial in two samples of 48 and 32 participants found that both single doses (up to 600 mg) and repeated ascending doses (up to 450 mg twice /day) of GET-73 were safe and well tolerated [250]. Another inpatient human laboratory study conducted by the same group confirmed the safety and tolerability of GET73 such that no serious or severe adverse events occurred when GET73 was co-administered with alcohol. Co-administration also did not affect the pharmacokinetics of either GET73 or alcohol. However, GET73 also had no effect on alcohol cue-induced craving or self-administration, warranting additional research [251]. A clinical trial of the effects of GET73 on magnetic resonance spectroscopy (MRS) measures of glutamate and GABA levels in individuals with AUD was recently completed (NCT03418623); however, the results have not yet been posted, and another trial, which includes a free-drinking bar lab component, is ongoing (NCT04831684).

In comparison to the drawbacks presented by GABA_B agonists like baclofen and SMO, positive allosteric modulators (PAMs) at this receptor present a potential alternative [252]. By binding to a different, non-competitive allosteric site on the GABA_B receptor, PAMs allow endogenous GABA, binding at its original orthosteric site, to retain its potency and efficacy, reducing the risk of tolerance development and side effects [253, 254]. A number of PAMs have been studied as potential pharmacotherapies for AUD, and have demonstrated reductions in alcohol-associated behaviors and ethanol self-administration in preclinical models [252, 255‐258]. When directly contrasted against baclofen, a GABA_B PAM had a better profile, with dose-dependent reduction of relapse-like alcohol drinking and with no signs of sedation [257].

One such novel GABA_B PAM, ASP8062, appears particularly promising as it has been shown to significantly increase the affinity and efficacy of endogenous GABA binding in human and rat GABA_B receptors in vitro and with oral administration in an in vivo rodent model of fibromyalgia, demonstrating the ability of oral formulated ASP8062 to cross the blood-brain-barrier [259]. ASP8062 has recently progressed to clinical development. Two Phase I clinical trials with a combined total of 112 participants evaluated single ascending doses and multiple ascending doses of ASP8062, respectively [260]. These studies found that ASP8062 was well tolerated in humans, with no evidence of drug effects on safety, cognition, drug withdrawal, or suicidal ideation. One additional clinical trial, assessing the safety and efficacy of ASP8062 for alcohol use disorder (NCT05096117), is currently underway. Overall, ASP8062, and GABA_B PAMs in general, appear to be well tolerated in humans and decrease alcohol self-administration in animals. These agents present a potential pathway to better utilize the GABAergic system and reduce the side effects seen with GABA_B agonists.

Ghrelin, a peptide produced by endocrine cells primarily in the stomach [261], is thought to regulate growth hormone secretion, food intake, and glucose homeostasis [262]. Ghrelin is also thought to play a role in AUD. Ghrelin signaling is required for stimulation of the reward system by alcohol, and higher ghrelin levels are associated with higher self-reported measures of alcohol craving [263]. In human laboratory studies, intravenous ghrelin administration has been shown to increase the urge to drink, increase alcohol self-administration, and modulate brain activity in regions involved in reward processing and stress regulation [264, 265].

Preclinical studies with ghrelin receptor (GHS-R1a) antagonists have shown reductions in alcohol conditioned place preference, alcohol intake and preference, and alcohol-elicited nucleus accumbens dopamine release in rodents [266‐270].

PF-5190457 is a ghrelin receptor inverse agonist that inhibits GHS-R1a constitutive activity as well as blocking its activation by ghrelin [271]. In a preliminary clinical study in 12 heavy-drinking individuals, PF-5190457, compared to placebo, reduced alcohol craving and cue-reactivity to alcohol. Additionally, when administered in combination with alcohol, PF-5190457 was safe and well-tolerated with no drug-alcohol interactions [272]. This was the first clinical study of a GHS-R1a inverse agonist in a sample of heavy alcohol drinkers.

PF-5190457 may increase somnolence, heart rate, and lower blood glucose concentrations [273], although clinical results indicate that these side effects were not exacerbated by alcohol co-administration and in general PF-5190457 is well tolerated. In summary, preclinical and early clinical evidence support additional research toward investigating PF-5190457 as a pharmacological approach to treat AUD.

Cannabidiol (CBD), one of the main compounds found in Cannabis sativa, has shown promise as a novel therapeutic to treat AUD. CBD is nonintoxicating and has diverse pharmacological effects throughout the central nervous system, including functioning as a negative allosteric modulator of CB₁ and CB₂ receptors [274, 275], and blocking anandamide update and inhibiting enzymatic hydrolysis [276]. CBD may also interact with non-endocannabinoid signaling systems, including the serotonergic system [277] and the opioidergic system [278], among others.

Preclinical studies have shown that CBD reduces alcohol administration, decreases motivation for alcohol, reduces relapse-like behavior, and improves withdrawal symptoms in animals exposed to chronic alcohol [279]. Evidence in healthy individuals demonstrates that CBD is well tolerated, does not interact with the subjective effects of alcohol, and has no abuse liability [280]. Two recent studies investigated signals for potential efficacy of CBD as a treatment for heroin use disorder [281] and cannabis use disorder [282]. Regarding heroin use disorder, acute CBD reduced cue-induced craving for heroin and reduced anxiety in a sample of 42 abstinent individuals, which was maintained one-week following the last CBD exposure. The cannabis use disorder study found that CBD was more efficacious at reducing cannabis use than placebo in a sample of 48 subjects. In both clinical samples, CBD was well tolerated and not associated with serious adverse events. CBD is currently being evaluated as a potential treatment for AUD, AUD with comorbid PTSD, and alcohol withdrawal in AUD in three clinical trials (NCT03252756, NCT03248167, NCT04205682). In brief, preclinical evidence and clinical evidence in other substance use disorders indicate the promise of CBD as a novel therapeutic for AUD.