Initial enthusiasm for the use of naltrexone to maintain sobriety in patients with alcohol dependence has been complicated by inconsistent results from smaller studies conducted at single trial sites. Krystal and colleagues conducted a multi-center, randomized, placebo-controlled study of naltrexone administered as an adjunct to psychosocial treatment in 627 patients with alcohol dependence.

Study participants were enrolled from Veterans Affairs outpatient alcohol treatment programs. After five days of sobriety, study participants were randomized to receive either 12 months of naltrexone (25 mg once daily for two days, then 50 mg once daily for three or 12 months), placebo for 12 months, or naltrexone for three months followed by placebo for nine months. All participants received individual psychotherapy for 13 months and were encouraged to attend Alcoholics Anonymous meetings. Compliance with medication was monitored by use of electronic bottle caps that recorded the date and time of each opening and that showed the number of hours that had elapsed since the previous opening. Almost all of the participants were male, and approximately 35 percent were from minority ethnic groups.

Overall, 73 percent of the participants completed the study, with no significant differences in dropout rates among the three treatment groups. Pain, flu-like symptoms, and headache were the most commonly reported adverse events occurring with use of naltrexone, but the rates of these symptoms did not differ significantly from placebo rates.

Neither short- nor long-term use of naltrexone significantly improved the ability of participants to maintain sobriety. At 13 weeks, relapse in drinking occurred in 44 percent of participants taking placebo versus 38 percent of those taking naltrexone. The total number of days when drinking occurred and the number of drinks per day in participants who relapsed was not significantly improved in the groups taking naltrexone. Compliance with medication was poor in all groups by completion of the study (42 to 44 percent of total doses were taken among the three groups).

The authors concluded that naltrexone combined with psychosocial therapy was not effective in treating men with chronic, severe alcohol dependence.