Alcohol use disorder: Pharmacologic management

INTRODUCTION — 

Alcohol use disorders are among the most prevalent of all substance use disorders worldwide. The global single-year prevalence has been estimated to be over 100 million individuals [1]. Additionally, nearly 3 million deaths (5.3 percent of all deaths globally) have been attributed to alcohol in a single year [2].

Pharmacologic treatment of alcohol use disorder has focused on altering the reinforcing effects of alcohol use. Medication development has focused on several neurotransmitter systems that mediate reinforcement including opioid, glutamate, gamma-aminobutyric acid, and serotonin systems.

Several medications can be used to treat alcohol use disorder, leading to reduced heavy drinking and increased days of abstinence [3]. These outcomes likely reduce the overall risk associated with alcohol use disorder despite total abstinence not being achieved.

This topic reviews the pharmacotherapy for treatment of alcohol use disorder. The epidemiology, pathogenesis, clinical manifestations, assessment, and diagnosis of risky drinking and alcohol use disorder, as well as psychosocial treatments for the disorder, are discussed separately. Other topics related to alcohol use disorder including the management of withdrawal and nutritional issues are also reviewed separately.

●(See "Risky drinking and alcohol use disorder: Epidemiology, clinical features, adverse consequences, screening, and assessment".)

●(See "Screening for unhealthy use of alcohol and other drugs in primary care".)

●(See "Brief intervention for unhealthy alcohol and other drug use: Efficacy, adverse effects, and administration".)

●(See "Alcohol withdrawal: Ambulatory management".)

●(See "Management of moderate and severe alcohol withdrawal syndromes".)

WHO WE TREAT — 

Pharmacotherapy is a component of the treatment of alcohol use disorder that is often combined with psychosocial interventions. The decision to treat alcohol use disorder with medication management is based on the severity of the disorder; however, patient preference is an important factor as all treatment decisions are made by shared decision making. In general, we include medications, along with psychosocial intervention as part of the treatment approach in patients with moderate to severe alcohol use disorder, or in those with mild disorder who request it. Pharmacologic management has not been extensively studied in individuals with mild alcohol use disorder. A discussion of the diagnostic criteria for alcohol use disorder including severity specifiers is found on the table and discussed elsewhere (table 1). (See "Alcohol use disorder: Treatment overview" and "Risky drinking and alcohol use disorder: Epidemiology, clinical features, adverse consequences, screening, and assessment".)

CHOOSING INITIAL AGENT — 

Several agents are effective in the treatment of alcohol use disorder. However, minimal direct evidence supports one treatment as compared with another, therefore our choice is based on other factors such as prior history, presence of co-occurring conditions, treatment goal, and patient preference. In patients who have responded to a prior medication, we typically choose that medication again [4-16]. Our preference for choosing pharmacotherapy for individuals with and without co-occurring opioid use disorder is found in the associated algorithms (table 2 and algorithm 1 and algorithm 2).

Patients without co-occurring disorders — For individuals without co-occurring disorders the choice of agent is guided by treatment goal and patient preferences:

●Treatment goal is complete abstinence – In individuals whose goal of treatment is complete abstinence, we typically use disulfiram. This is particularly true in an individual who is able to take the medication with supervision. However, we avoid disulfiram in those with advanced liver disease or in those with difficulty understanding the interaction between disulfiram and alcohol (eg, disulfiram reaction). Disulfiram is typically avoided in suicidal patients and may be ill-advised in very impulsive patients who have previously ingested significant amounts of alcohol while receiving this medication. (See 'Considerations for specific comorbidities' below.)

●Treatment goal is reduction of use – In individuals whose goal is reduction of use we choose from among naltrexone, acamprosate, and topiramate. Patient preference is a strong consideration (eg, monthly injection versus daily medication). The presence of certain co-occurring disorders makes certain options more or less preferable, as discussed below. (See 'Considerations for specific comorbidities' below.)

Considerations for specific comorbidities — The presence of specific co-occurring disorders is a prominent factor in our choice of initial pharmacologic management of alcohol use disorder. In individuals with one or more co-occurring disorders our choice of pharmacologic management is dependent on the effects of the disorder on the agent (eg, liver dysfunction impairs metabolism of agents) and the potential benefits of the agent in treating the co-occurring disorder (eg, topiramate in appropriate individuals with a seizure disorder, naltrexone in individuals with opioid use disorder). (See 'Opioid use disorder or prescribed opioids' below and 'Advanced liver disease' below and 'Seizure disorder' below.)

Opioid use disorder or prescribed opioids

●Co-occurring opioid use disorder – For patients with alcohol use disorder and co-occurring opioid use disorder, we prefer to treat both disorders with naltrexone (algorithm 1). However, naltrexone can only be started after a sufficient time has elapsed since last opioid exposure and cannot be used in individuals treated with opioid agonist for opioid use disorder (eg, buprenorphine, methadone). In these cases, we would use acamprosate, topiramate, or disulfiram; the choice among these treatments is based on the presence of other co-occurring disorders, goal of treatment and patient preference. (See 'Choosing initial agent' above.)

Naltrexone administration, dosing, and adverse effects in the treatment of substance use disorders is discussed below and elsewhere. (See "Opioid use disorder: Pharmacologic management", section on 'Naltrexone: Opioid antagonist' and 'Naltrexone' below.)

●Clinically indicated opioid use – For patients with alcohol use disorder who are taking opioid agonists (eg, oxycodone) for a clinically indicated use (eg, pain management), naltrexone must be avoided. (See "Use of opioids in the management of chronic pain in adults".)

In these cases, we would use acamprosate, topiramate, or disulfiram and base the choice on presence of other co-occurring disorders, prior history, goal of treatment and patient preference [16]. (See 'Choosing initial agent' above.)

Advanced liver disease — The presence of advanced liver disease affects our choice of medication for alcohol use disorder:

●For Child-Pugh Class C (decompensated cirrhosis), we typically choose acamprosate as it is metabolized primarily through the kidneys. If acamprosate is ineffective or unavailable, topiramate or naltrexone may be used with extreme caution. Disulfiram is avoided.

●For Child Pugh Class B (significant functional compromise), acamprosate is our first choice; however, topiramate and naltrexone are reasonable alternatives. We generally avoid disulfiram; however, it can be used with caution if necessary.

●For Child Pugh Class A (compensated cirrhosis), our choice is based on other factors (eg, prior history, preference, goals of treatment). Child Pugh Class A does not impact our choice of drug.

In patients with advanced liver disease, baclofen is another option that appears to be effective in some trials. (See 'Second-line agents' below and "Cirrhosis in adults: Overview of complications, general management, and prognosis", section on 'Child-Pugh classification'.)

Seizure disorder — Topiramate is our preferred choice of medications in patients who have a co-occurring seizure disorder that would appropriately be treated with this agent [16].

If topiramate is not an option (stable on another antiseizure medication, prior history of nonresponse) we would use naltrexone or acamprosate depending on other factors such as patient preference.

In individuals with a seizure disorder, we avoid disulfiram due to the possibility of intractable nausea and vomiting associated with the disulfiram-alcohol reaction. Intractable vomiting may lead to electrolyte abnormalities which may lower the seizure threshold.

Kidney function impairment — We are cautious when prescribing acamprosate to individuals with kidney dysfunction. Acamprosate is contraindicated in individuals with creatinine clearance ≤30 mL/min.

Cognitive dysfunction — We avoid topiramate in individuals with cognitive dysfunction as topiramate is associated with impaired cognition. (See "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects", section on 'Topiramate'.)

We avoid disulfiram if there are concerns about the patient’s capacity to understand the implications of consuming alcohol while taking disulfiram.

Special considerations

Pregnancy — We refer pregnant individuals with a substance use disorder to a clinician with expertise in treating substance use disorders in pregnancy.

Psychosocial treatments are prioritized as there is a paucity of data on the safety of pharmacologic therapies for alcohol use disorder in pregnant individuals [12]. (See "Alcohol intake and pregnancy", section on 'Management of screen-positive pregnant persons'.)

If abstinence is not achieved without the use of medications after four to six weeks, the risks of continued heavy drinking likely outweigh the possible adverse effects of medication. In weighing risks and benefits of prospective treatment, one should consider potentially harmful effects of alcohol to the mother and to the developing fetus, with alcohol a known teratogen and the most common cause of congenital anomaly in the United States.

Selection of a specific agent is guided by limited data. One small study suggested no clear association between exposure to acamprosate with poor maternal or neonatal health outcomes [17]. Acamprosate may also be favored because opioids may be desired around delivery. However, naltrexone has been used more widely for substance use disorder during pregnancy [18]. When naltrexone is used, we discontinue the medication a few weeks before the expected delivery date.

Topics related to substance use, including alcohol use during pregnancy, are reviewed separately [19,20]. (See "Substance use during pregnancy: Screening and prenatal care" and "Alcohol intake and pregnancy".)

Patients hospitalized for alcohol-related disorders — Medication for alcohol is underused, despite the morbidity and mortality associated with alcohol use disorder, the known efficacy of medications for alcohol use disorder, and the guideline recommendations [21].

Hospitalization may provide an opportunity for initiating medication for alcohol use disorder. Initiating pharmacologic management in individuals hospitalized for alcohol-related disorders (eg, alcohol-related gastritis, alcohol-related cardiomyopathy, pancreatitis, liver disease) appears to improve outcomes. In an observational study including over 6700 individuals hospitalized at least once for alcohol-related disorders, initiation of medication for alcohol use disorder prior to discharge is associated with better outcomes such as all-cause mortality, all-cause emergency department visit, or all-cause readmission (composite adjusted relative risk 0.58, 95% CI 0.45-0.76) [22]. Additionally, at 30-day follow-up, initiation of medication for alcohol use disorder led to lower rate of alcohol-related return to emergency department or hospitalization (relative risk 0.49, 95% CI 0.34-0.71).

Our approach to choosing initial management of these individuals is the same as for the general population. (See 'Choosing initial agent' above.)

Dose, efficacy, adverse effects of initial pharmacotherapies

Naltrexone — Naltrexone is an effective treatment in the management of alcohol use disorder [4,5,7-9,23]. The preferable dosing schedule, its ability to be used in individuals who are still drinking, and its efficacy in individuals with co-occurring opioid use disorder make it an attractive agent for many.

Naltrexone exerts its principal pharmacologic effects through blockade of the mu-opioid receptor. Endogenous opioids modulate alcohol’s reinforcing effects [24,25]. Naltrexone also modifies the hypothalamic-pituitary-adrenal axis to suppress ethanol consumption [26].

●Formulations – Naltrexone can be taken orally or administered in a long-acting injectable (LAI) form. Both oral and long-acting dosing have demonstrated efficacy compared with placebo. Comparative studies are limited and have come to different conclusions regarding comparative efficacy [27-29].

For most patients treated with medication for alcohol use disorder, the decision to begin oral versus intramuscular (IM) is based on patient preference. We typically prefer starting IM naltrexone to ensure adherence [30]. However, some may prefer to start with oral medication to see if side effects emerge or liver enzymes are affected, before committing to a longer course of treatment. Depot preparations of naltrexone may improve adherence by reducing the frequency of medication administration from daily to monthly. However, they do require an injection visit. Some patients may adhere better to daily medication whereas others may be willing to attend monthly visits. Additionally, a steady state of medication level is achieved with LAI naltrexone. This may avoid peak effects that might exacerbate adverse events [31].

•LAI naltrexone – LAI naltrexone is given as an IM injection of 380 mg every four weeks to the gluteal area. Common adverse events observed among individuals receiving LAI naltrexone included nausea (33 percent), fatigue (20 percent), and decreased appetite (13 percent).

The US Food and Drug Administration (FDA) has reported 196 cases of serious injection site reactions from postmarketing surveillance including induration, cellulitis, hematoma, abscess, and necrosis. Females appear to be at higher risk for this reaction. Patients should report injection-site pain, swelling, bruising, pruritus, or redness, and seek medical attention if symptoms are not improving after one week. Liver enzymes should be monitored within several weeks of initiating treatment and then every six months during ongoing treatment.

LAI naltrexone appears to be superior to placebo in reducing drinking and heavy drinking among adults with alcohol use disorder [4,7,8]. As examples:

-In a meta-analysis, individuals with alcohol use disorder treated with LAI naltrexone had fewer days of drinking per month compared with individuals receiving placebo (five trials, n = 314; weighted mean difference -2, 95% CI -3.39 to -0.61) [8]. Additionally, individuals treated with LAI naltrexone had fewer heavy drinking days per month (defined as four or more drinks/day in females; five or more drinks/day in males) than those in placebo group (seven trials, n = 881; weighted mean difference -1.2, 95% CI -2.1 to -0.23).

-In a meta-analysis of two trials investigating treatment for alcohol use disorder, treatment with LAI naltrexone, as compared with placebo, was associated with a greater reduction in percentage drinking days (two trials, n = 467; weighted mean difference -5, 95% CI -9.5 to -0.5) and heavy drinking days (three trials, n = 956; weighted mean difference -4.7, 95% CI -8.6 to -0.73). However, LAI naltrexone, as compared with placebo, was not associated with lower rates of return to drinking (two trials, n = 939) or return to heavy drinking (two trials, n = 615) [4].

•Oral naltrexone – For patients who choose oral naltrexone, our practice is to start oral naltrexone at 50 mg per day and monitor for side effects. However, to avoid side effects, some practitioners start at 25 mg and titrate after several days. Doses up to 100 mg have been effective in clinical trials [32].

Targeted (as needed) dosing may be a useful strategy for promoting adherence in young individuals who often prefer to take medication on an as needed basis [33,34]. While we typically prescribe naltrexone for daily use, targeted dosing may be an effective alternative or supplement to daily dosing. In one trial including 140 young adults with heavy drinking (ie, four or more heavy drinking days over the past four weeks), the use of targeted naltrexone was found to reduce the likelihood of intoxication by nearly 23 percent versus placebo [33,34].

Meta-analyses of clinical trials for alcohol use disorder have found oral naltrexone to reduce alcohol consumption compared with placebo [4,9,23]. As an example, in a meta-analyses of treatment for alcohol use disorder, treatment with oral naltrexone 50 mg, as compared with placebo, was associated with decreased return to any drinking (16 trials, n = 2347; relative risk 0.93, 95% CI 0.87-0.99), decreased return to heavy drinking (23 trials, n = 3170; relative risk 0.81, 95% CI 0.72-0.9), decreased percentage of drinking days (15 trials, n = 1992; weighted mean difference -5.1, CI -7.2 to -3.0) and percentage of heavy drinking days (seven trials, n = 624; weighted mean difference -4.3, CI -7.6 to -0.9) [4]. These finding support those of a prior meta-analysis [9].

●Adverse effects – Side effects of naltrexone are nausea, headache, and dizziness, which subside with continued use.

We monitor liver enzymes within several weeks of initiating treatment and then every six months during ongoing treatment. Fivefold elevation in liver enzymes occurred in 11 of 614 individuals who received naltrexone in the Combined Pharmacotherapies and Behavioral Interventions (COMBINE) study [32]. Enzyme levels returned to baseline after discontinuation of medication in the nine individuals who had stopped drinking and returned for follow-up. Elevations in liver enzymes are therefore largely attributable to heavy drinking; in a large safety study (865 patients), liver enzymes did not differ from a comparison group not taking naltrexone [35].

●Contraindications – Naltrexone should be avoided in individuals using opioids or prescribed opioids for pain management. Caution is advised when treating patients with acute hepatitis or hepatic failure, though naltrexone appears to be safe and effective in patients with Child-Pugh class A or B cirrhosis [36-38]. (See 'Advanced liver disease' above.)

Acamprosate — Acamprosate's principal antidrinking neurochemical effect has been attributed to the modulation of glutamate neurotransmission at metabotropic-5 glutamate receptors [39].

●Administration – We begin treatment with acamprosate when abstinence is achieved (eg, after medically supervised withdrawal) and, typically, maintain treatment during return to use. The usual dose for acamprosate is 666 mg three times daily. However, in individuals with moderate kidney dysfunction (creatinine clearance 30 to 50 mL/min) an initial dose of 333 mg three times daily is recommended. Additionally, in individuals with a body weight <60 kg we typically initiate treatment at a lower dose (eg, 333 mg twice daily).

●Efficacy – Meta-analyses have shown that acamprosate is effective in maintaining abstinence in individuals with alcohol use disorder who were recently withdrawn from alcohol use [4-6]. As examples:

•In a meta-analysis of 18 studies including nearly 2300 individuals who were recently withdrawn from alcohol use, acamprosate increased the likelihood of continuous abstinence versus placebo at 12 months (odds ratio 1.86, 95% CI 1.49-2.33) [6].

•In another meta-analysis investigating pharmacologic treatments for alcohol use disorder, treatment with acamprosate, as compared with placebo was associated with decreased return to any drinking (20 trials, n = 6380; relative risk 0.88, 95% CI 0.83-0.93), and percentage of drinking days (14 trials, n = 4916; weighted mean difference -8.3, 95% CI -12.2 to -4.4) [4].

●Adverse effects – The most prominent adverse effects of treatment with acamprosate include diarrhea, nervousness, and fatigue. These usually subside with continued use. Because acamprosate is excreted mostly unchanged by the kidneys, rather than the liver, it can be used safely in individuals with alcohol use disorder and liver disease.

●Contraindications – Acamprosate is contraindicated in individuals with severe kidney dysfunction (creatinine clearance ≤30 mL/min).

Disulfiram — Disulfiram is an aversive agent that does not directly influence motivation to drink but discourages drinking via the threat of the “disulfiram reaction.” By blocking a key step in the enzymatic breakdown of alcohol, disulfiram causes an accumulation of alcohol’s primary metabolite, acetaldehyde. Within 10 to 30 minutes of consuming alcohol, this accumulation causes unpleasant effects such as sweating, headache, dyspnea, lowered blood pressure, flushing, sympathetic overactivity, palpitations, nausea, and vomiting [40].

●Administration – Disulfiram is initially given at 250 to 500 mg/day for one to two weeks, followed by an average maintenance dose of 250 mg/day with a range from 125 to 500 mg based on the severity of adverse effects.

Forty-eight hours of total abstinence is needed prior to starting disulfiram. Patient education should address "hidden" forms of ethanol (eg, tonics and mouthwashes) and also the duration of the drug's activity (up to 14 days after stopping).

Supervised dosing is not required for this medication, as highly motivated patients likely do not require this level of oversight. However, prior research has suggested that patients who do have the benefit of supervised disulfiram dosing, via strong social support, tend to fare substantially better than controls with regards to sustained abstinence [13].

●Efficacy – While several double-blinded randomized controlled trials have failed to demonstrate a treatment effect for disulfiram [13], many investigators have argued that the unique deterrent effect of disulfiram necessitates open-labeled studies to investigate its efficacy [13-15,41,42].

•In a meta-analysis of 17 open-labeled randomized controlled trials treatment with disulfiram led to a medium to large treatment effect as compared with controls (which included both placebo and medication comparators [naltrexone and acamprosate]) on abstinence outcomes (Hedges’ g = 0.7, 95% CI 0.46-0.93) [13].

•In a network meta-analysis of 156 randomized controlled trials with over 27,000 participants, supervised treatment with disulfiram was among the most effective medications. In comparison with placebo, disulfiram was effective in maintaining abstinence (relative risk 1.71, 95% CI 1.39–2.10) and reducing heavy drinking (relative risk 0.19, 95% CI, 0.10–0.35) [43].

●Disulfiram reaction – Reactions to disulfiram are self-limited. However, some individuals taking disulfiram who drink alcohol or alcohol-containing products can have a severe reaction to the combination. The severity and duration of the reaction depends on the amount of alcohol ingested. The reaction may last for several hours or up to a day.

Individuals may present with symptoms such as chest pain, confusion, headache, and vomiting that require further evaluation for myocardial infarction and other etiologies. Once myocardial infarction is excluded in patients with chest pain, treatment is primarily supportive (antiemetics for vomiting; Trendelenburg positioning and intravenous fluids for orthostatic hypotension; diphenhydramine for flushing).

Fomepizole has been suggested for life-threatening symptoms of acetaldehyde in patients with severe presentations. Fomepizole (4-methylpyrazole), administered as a single intravenous 7.5 mg/kg dose, blocks alcohol dehydrogenase and has been shown to reverse disulfiram reactions in a small case series of patients with nonspecific electrocardiogram changes with chest pain who are unresponsive to fluid administration and norepinephrine [44,45].

●Contraindications – Contraindications to disulfiram include Child-Pugh class B or C cirrhosis, clinically significant coronary artery disease, psychosis and known hypersensitivity to the medication or other thiuram derivatives. We do not use disulfiram in individuals with a seizure disorder due to the potential of a disulfiram reaction which may cause intractable nausea and vomiting and metabolic dysregulation [46,47].

●Adverse effects – Side effects of disulfiram are usually minor, including fatigue, mild drowsiness, headache, and dermatitis. Severe adverse reactions are rare but include psychosis and hepatitis. Individuals receiving disulfiram should be monitored for hepatotoxicity several weeks after initiating treatment and then every six months if treatment with disulfiram continues [48]. It is also important to educate patients about possible signs and symptoms of liver injury including nausea and vomiting, abdominal pain, malaise, fever, and jaundice and to seek medical attention if such events occur.

Topiramate — Topiramate, an anticonvulsant medication with pharmacologic properties including blocking of voltage-dependent sodium channels, potentiation of gamma-aminobutyric acid mediated transmission and antagonism of glutamate receptors, has been found to decrease alcohol use in individuals with alcohol use disorder.

●Administration – Topiramate is initiated at a dose of 25 mg daily and can be slowly titrated up to a maximum dose of 300 mg/day, over eight weeks. A titration schedule is shown in the table (table 3).

●Efficacy – In clinical trials, topiramate reduces alcohol consumption compared with placebo [9-11]. In a meta-analysis of seven randomized trials including 1125 individuals with alcohol use disorder, treatment with topiramate, as compared with placebo, demonstrated higher rates of abstinence and lower rates of heavy drinking, but similar measure of alcohol craving efficacy [10]. The specific outcome measures used varied across studies, but overall, the effects were judged to be small to moderate.

Furthermore, in one randomized trial including 147 individuals assigned to 12 weeks of treatment with topiramate versus naltrexone, topiramate appeared to be as effective and safe as naltrexone on self-reported measures of alcohol consumption. On some measures (eg, standard drinks per day, heavy drinking days) topiramate appeared to have greater effects [49].

●Adverse effects – Adverse effects associated with topiramate include cognitive impairment (eg, word-finding difficulties), paresthesias, weight loss, headache, fatigue, dizziness, and depression. To some individuals many of these are intolerable; however, topiramate generally appears to be well tolerated with equivocal data suggesting higher dropout rates for patients taking topiramate as compared with placebo [10]. (See "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects", section on 'Topiramate'.)

MONITORING AND DEFINING RESPONSE — 

Monitoring of individuals with alcohol use disorder is discussed elsewhere. Response to treatment is based on the goals of treatment and is also discussed elsewhere. (See "Alcohol use disorder: Treatment overview", section on 'Monitoring and evaluating response'.)

SUBSEQUENT TREATMENT

Good response to initial agent — We continue individuals with alcohol use disorder who have met their treatment goal (eg, continued abstinence, adequate reduction of use) with pharmacologic management for at least one year as this duration of time is associated with a lower risk of recurrence [50]. When discontinuing medication, we taper patients off slowly (eg, over several weeks). (See "Alcohol use disorder: Treatment overview", section on 'Patients with robust response or in remission' and "Substance use disorders: Continuing care treatment", section on 'Duration of treatment'.)

Inadequate response to initial agent — For individuals who do not respond to initial pharmacologic management after several months (eg, four to six months) with one of the agents listed above, our preference is to discontinue the first agent and try subsequent trials of each of the other suggested initial pharmacotherapies, as clinically indicated. (See 'Dose, efficacy, adverse effects of initial pharmacotherapies' above.)

We use the same factors (co-occurring disorders, past history, treatment goal) that we used for the first agent in choosing subsequent agents. Treatment decisions are made using shared decision making. We do not combine or use supratherapeutic dosing. There are no consistent data suggesting that combination pharmacotherapy offers any advantages over monotherapy. Furthermore, trials have not supported using higher doses of medication than suggested. [32,51,52].

Additionally in all individuals with inadequate response to initial medication we encourage psychosocial treatment if not already engaged in this. If engaged in psychosocial treatment we assess for fidelity of treatment (eg, attending treatment, engaged and motivated to work in treatment). We increase the intensity (eg, frequency) of the psychosocial treatment or add another psychosocial intervention as a next step. (See "Alcohol use disorder: Treatment overview", section on 'Psychosocial strategies' and "Alcohol use disorder: Psychosocial management", section on 'Interventions'.)

SECOND-LINE AGENTS — 

For individuals who have not responded adequately to each of the available initial agents our preference is a trial of gabapentin or baclofen. (See 'Dose, efficacy, adverse effects of initial pharmacotherapies' above.)

●Gabapentin – Gabapentin may be a reasonable option for patients who may have not successfully reached their treatment goal despite a trial of each of the indicated medications above. However, one consideration is the addictive potential of gabapentin.

Clinical trials testing the efficacy of gabapentin at various doses from 900 to 3600 mg have found mixed results in treatment for alcohol use disorder [53-55]. In a meta-analysis of seven placebo-controlled randomized controlled trials, gabapentin was effective only on a single drinking outcome – percent of heavy drinking days (g = 0.64, 95% CI 0.06-1.22) [56]. In one small trial, the efficacy of gabapentin in reducing heavy drinking appeared more pronounced in those with withdrawal symptoms compared with those without [55]. (See "Alcohol withdrawal: Ambulatory management", section on 'Very mild withdrawal (CIWA-Ar <10)'.)

●Baclofen – Baclofen is another option for treatment that appears to be effective in some, but not all, trials [4,43,57-63]. Some providers prefer baclofen due to its kidney, rather than hepatic metabolism. (See 'Advanced liver disease' above.)

THERAPIES WITH UNCLEAR EFFICACY — 

Other medications with little data to support their use are discussed below. Due to limited empirical support in the treatment of alcohol use disorder, we cannot recommend them as first- or second-line agents until further studies support their efficacy.

●Ketamine – There has been a resurgence of interest in the role of ketamine, a decades-old dissociative anesthetic, in the treatment of several substance use disorders, including alcohol use disorder. In a review including seven studies using ketamine for the treatment of alcohol use disorder, ketamine, typically administered by intravenous infusion, was found to reduce heavy drinking days and promote abstinence relative to usual care. However, additional studies are needed before ketamine can be recommended for alcohol use disorder [64].

●Glucagon-like peptide 1 (GLP-1) receptor agonists – Initial randomized trials have reported mixed results regarding the efficacy of GLP-1 receptor agonists in treating alcohol use disorder. In a phase 2 placebo-controlled, double-blind trial of 48 participants with alcohol use disorder, subcutaneous semaglutide at doses from 0.25 mg to 1 mg weekly reduced weekly alcohol cravings and the number of heavy drinking days (approximately one day less per week) at nine weeks, compared with placebo [65]. By contrast, a placebo-controlled, double-blind trial of exenatide 2 mg weekly did not reduce heavy drinking days in 127 participants with alcohol use disorder [66]. Data from cohort studies and secondary analyses of randomized trials [67] suggest that GLP-1 agonists may have efficacy in reducing alcohol use [68] and alcohol-related complications, such as hospitalizations [69], in individuals who are taking these medications for other indications. These results suggest the need for larger randomized trials.

●Nalmefene – Nalmefene, an opioid antagonist, has been found to reduce drinking in individuals with alcohol use disorder using a targeted dosing strategy; however, the methodologic rigor of the trials have been questioned [70,71]. Nalmefene has several potential advantages over naltrexone, including absence of dose-dependent liver toxicity, longer-acting effects, and more effective binding to central opiate receptors. Nalmefene is approved for treatment of alcohol use disorder in the European Union. It is available in the United States as treatment for opioid overdose [9,72-75].

●Selective serotonin reuptake inhibitors (SSRIs) – Preliminary evidence suggests that subtypes of alcohol dependence may respond differently to serotonergic drugs, with more favorable outcomes seen in the group characterized by a later age of onset, less psychosocial morbidity, and low familial loading. Additionally, SSRIs may be useful in treating individuals with depression and substance use disorder, commonly occurring comorbidities [76-79].

●Ondansetron – This is a serotonin 5-HT3 receptor antagonist used to treat chemotherapy-induced nausea. Ondansetron may be selectively efficacious in individuals with early-onset subtype of alcohol use disorder (onset of problem drinking prior to age 25 years) and in individuals with family history of alcohol use disorder [80-82].

●Varenicline – While an earlier study reported on the potential for varenicline to reduce alcohol consumption in individuals with alcohol use disorder [83], this effect might be more limited [84] and in need of further validation. Some studies suggest that varenicline treatment may be associated with reduced drinking in patients with alcohol addiction who smoke and in patients with alcohol use disorder and depression [85,86].

●Psilocybin – Psilocybin administered in combination with psychotherapy appears to decrease the number of heavy drinking days in individuals with alcohol use disorder. In a trial, 95 subjects with alcohol use disorder were randomly assigned to two day-long medication sessions (weeks 4 and 8) with psilocybin or diphenhydramine (active control), each in addition to 12 weeks of manualized psychotherapy [87]. Over the 32-week follow-up period, subjects in the psilocybin group reported a lower percentage of heavy drinking days than those in the diphenhydramine group (9.7 versus 23.6, respectively; mean difference 13.86, 95% CI 3.0-24.7). Furthermore, subjects in the psilocybin group had fewer mean drinks per day than the diphenhydramine group (1.2 versus 2.3, respectively; mean difference 1.1, 95% CI 0.27-0.92). No serious adverse events were reported among either group. Although psilocybin is classified as a Schedule I controlled substance in the United States (no accepted medical use and high potential for abuse), these data suggest that further evaluation of psilocybin for alcohol use disorder management may be warranted.

●Others – Prazosin and doxazosin (alpha-1 antagonists) [88] also have limited data to support their use.

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Alcohol use disorders and withdrawal".)

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, “The Basics” and “Beyond the Basics.” The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of individuals by searching on “patient info” and the keyword(s) of interest.)

●Basics topic (see "Patient education: Alcohol use – When is drinking a problem? (The Basics)")

●Beyond the Basics topic (see "Patient education: Alcohol use — when is drinking a problem? (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Choosing treatment modality – Pharmacotherapy is a component of the treatment of alcohol use disorder that is often combined with psychosocial interventions. In general, we include medications, along with psychosocial intervention as part of the treatment approach in patients with moderate to severe alcohol use disorder, or in those with mild disorder that request it (table 2). (See 'Who we treat' above.)

●Choosing initial agent – Several agents are effective in the treatment of alcohol use disorder. However, minimal direct evidence supports one treatment as compared with another. In patients who have responded to a prior medication, we typically choose that medication again (algorithm 1 and algorithm 2).

•Patients without co-occurring disorders – In an otherwise healthy individual without a prior history of treatment for alcohol use disorder, our choice is based primarily on treatment goal and is made by shared decision making.

If the goal of treatment is complete abstinence, we suggest disulfiram rather than other agents (Grade 2C).

If the goal of treatment is reduction of use, we choose from among naltrexone, acamprosate, and topiramate using shared decision making.

•Considerations for specific co-occurring disorders

-Opioid use disorder, prescribed opioids – For patients with alcohol use disorder and co-occurring opioid use disorder, our preference is to address both disorders with naltrexone (algorithm 1). However, naltrexone can only be started after a sufficient time has elapsed since last opioid exposure. (See 'Opioid use disorder or prescribed opioids' above.)

For patients with alcohol use disorder who are taking opioid agonists (eg, oxycodone) for a clinically indicated use (eg, pain management), naltrexone must be avoided. In these cases, we would use acamprosate, topiramate, or disulfiram and base the decision on other factors (eg patient preference, presence of other co-occurring disorders)

-Advanced liver disease – For decompensated cirrhosis (Child-Pugh Class C) Acamprosate is typically chosen as it is excreted unchanged primarily through the kidneys. Topiramate and naltrexone may be used with extreme caution. Disulfiram is avoided. (See 'Advanced liver disease' above.)

For child Pugh Class B (functional compromise) acamprosate is a reasonable choice. Topiramate and naltrexone are alternative choices. Disulfiram is used with caution. (See 'Advanced liver disease' above.)

For Child Pugh Class A (compensated cirrhosis) the disorder does not affect our choice. (See 'Advanced liver disease' above.)

-Seizure disorder – Topiramate is our preferred choice of medications in patients who have a co-occurring seizure disorder that would appropriately be treated this agent. We avoid disulfiram in individuals with a seizure disorder. (See 'Seizure disorder' above.)

•Other disorders or special considerations

-Kidney function impairment – We are cautious when prescribing acamprosate to individuals with kidney dysfunction. Acamprosate is contraindicated in individuals with creatinine clearance ≤30 mL/min. (See 'Kidney function impairment' above.)

-Cognitive dysfunction – We avoid topiramate in individuals with cognitive dysfunction. We avoid disulfiram in individuals with cognitive dysfunction if there are concerns about the patient capacity to understand its warning regarding use of alcohol. (See 'Cognitive dysfunction' above.)

-Pregnancy – We prefer psychosocial treatments, rather than medication management for pregnant individuals. There is a paucity of data on the safety of pharmacologic therapies for alcohol use disorder in pregnant individuals. (See 'Special considerations' above.)

●Subsequent treatment – For individuals with a good response to treatment (met treatment goals) with pharmacologic management, we continue treatment for at least one year. (See 'Good response to initial agent' above.)

For inadequate response to initial agent, we choose each of the other initial pharmacotherapy agents that are clinically acceptable in succession. We use the same factors (co-occurring disorders, history, patient preference) in choosing subsequent medication as we do for the first choice. Additionally, we add psychotherapy (if not already done), review for fidelity of psychotherapy and increase intensity of psychotherapy. (See "Alcohol use disorder: Psychosocial management" and 'Inadequate response to initial agent' above.)

For individuals with inadequate response to each of the clinically acceptable initial agents, we consider a trial of a second-line agent (ie, gabapentin or baclofen). (See 'Second-line agents' above.)

●Agents with unclear efficacy – While ketamine, glucagon-like peptide 1 agonists, nalmefene, and others have shown limited data supporting their use, we do not recommend these agents until further studies support their efficacy. (See 'Therapies with unclear efficacy' above.)

ACKNOWLEDGMENT — 

The UpToDate editorial staff acknowledges Bankole A Johnson, DSc, MD, MBChB, MPhil, FRCPsych, DFAPA, who contributed to earlier versions of this topic review.