Pharmacologic treatment of alcohol use disorder

Clinical question

Which Health Canada–approved pharmacologic treatments are effective for alcohol use disorder (AUD)?

Bottom line

Both acamprosate and naltrexone show benefit for alcohol abstinence compared with placebo. For every 12 patients treated with acamprosate, and every 20 patients treated with naltrexone, 1 fewer patient returns to drinking compared with placebo after 12 to 52 weeks. For harm reduction, naltrexone can reduce return to heavy drinking for 1 of every 13 patients.

Evidence

Results are statistically significant unless noted.

• In a systematic review of RCTs of 12- to 52-week treatments, most included supportive therapy and required detoxification. Results versus placebo were as follows¹:

  • -Return to any drinking:

    • — For acamprosate (16 RCTs, N = 4847), the most common dose was 666 mg 3 times per day. The rate was 76% versus 83% for placebo (number needed to treat [NNT] of 12).

    • — For 50 mg of oral naltrexone daily (16 RCTs, N = 2347), the rate was 63% versus 68% with placebo (NNT = 20).

    • — There was no difference with injectable naltrexone (2 RCTs, N = 939) or disulfiram (2 RCTs, N = 492).

  • -Return to heavy drinking:

  • — For 50 mg of oral naltrexone daily (19 RCTs, N = 2875), the rate was 46% versus 54% with placebo (NNT = 13).

  • — There was no difference with acamprosate (7 RCTs, N = 2496).

• Earlier systematic reviews of acamprosate² and naltrexone³ reported similar results.

• Evidence is insufficient or finds no benefit for acamprosate or naltrexone on mortality^1,4,5 or quality of life.¹

• The most common adverse effects for naltrexone^3,5 were nausea (26% vs 16% for placebo, number needed to harm [NNH] of 10) and sleepiness (21% vs 16% for placebo, NNH = 20). For acamprosate,⁴ the incidence of diarrhea (16% vs 10% for placebo, NNH = 17) decreases after the first 4 weeks of treatment.

Context

• Guidelines suggest that first-line pharmacotherapy include acamprosate for abstinence or naltrexone for reduced drinking or abstinence, and give practical prescribing tips.⁶

• Limited evidence has evaluated as-needed naltrexone. It might reduce alcohol consumption when used as cravings arise or before expected drinking.⁷

• Supportive interventions, including brief interventions in primary care, might benefit 1 in 10 individuals with excessive alcohol intake.⁸

• If patients do not respond to approved medications, a trial of alternative medications (eg, topiramate, gabapentin) might be reasonable.⁶

Implementation

In Canada, alcohol use is linked to about 8% of all deaths.⁶ Hospitalizations for AUD have increased over time, as has the number of patients presenting to primary care for alcohol-attributable diseases.⁹ Only limited numbers of individuals with AUD are offered treatment.⁶ The existence of multiple medications with evidence of benefit provides options for shared, informed decision making with patients. While abstinence is a laudable goal, reductions in return to heavy drinking or reduced heavy drinking days are also associated with improved outcomes.¹⁰