Alcohol abuse and alcoholism are a world-wide problem, from a medical but also a social viewpoint. In the past the therapy for alcoholics was mainly based on the psychosocial approach, but the combined use of pharmacotherapy and psychosocial interventions has raised the percentage rate of success in maintaining alcoholic patients in remission [1,2]. Currently there are only three medications that have been approved by the U.S. Food and Drug Administration (FDA) for use in treating alcohol abuse and alcoholism: disulfiram [3] naltrexone [4–6] and acamprosate [7,8]. Leading Italian studies also support the use of γ-hydroxybutyric acid (GHB). In alcohol dependence, in fact, one of the neurotransmitter systems most directly involved is the GABA-ergic one. γ-Aminobutyric acid (GABA) is the major inhibitory neurotransmitter in the central nervous system and this explains why GABA-ergic medications, such as GHB, are promising drugs with potential uses in treating alcohol withdrawal syndrome (AWS), but also in maintaining alcohol abstinence [9]. Since 1992, GHB has been approved in Italy and Austria as a treatment for alcohol dependence [10]. Many studies have focused on the efficacy of GHB in maintaining abstinence from alcohol [11–17]. The Cochrane Collaborative Review Group showed that GHB is effective in preventing relapses in previously detoxified alcoholics at a 3-month follow-up. The side-effects of GHB do not differ statistically from those induced by benzodiazepines, naltrexone or disulfiram [18]. A key drawback is that about 30–40% of alcoholic subjects are non-responders to GHB therapy [15]. In our clinical experience, the reduction, whether temporary or continuing, in the craving experienced by these patients was insufficient to control their dependence on alcohol. This could be related to the short half-life of GHB [19], so that it could be useful to divide up GHB administration into six doses per day [15,20]. Nevertheless, because of this last property some people may show a tendency to abuse GHB [21].

For over 55 years, disulfiram has been approved by the FDA for the treatment of alcohol dependence. It is a unique medication that relies on a ‘psychological threat’ to avoid disulfiram-ethanol reactions; in fact, it irreversibly inhibits aldehyde dehydrogenase (the enzyme that converts the relatively toxic metabolite acetaldehyde into the benign metabolite acetate), which is necessary to the metabolism of ethanol [22,23]. Consuming alcohol after taking disulfiram results in symptoms such as palpitations, flushing, nausea, vomiting, and headaches [24]. An evidence report from the Agency for Healthcare Research and Quality [3] concluded that four placebo-controlled randomized controlled trials (RCTs) using oral disulfiram produced mixed results. In two trials, oral disulfiram was shown to reduce the frequency of drinking days, but it failed to improve relapse rates compared with placebo [24]. Patients using disulfiram tended to remain in treatment longer than those who did not use disulfiram or used a naltrexone-based treatment [25,26]. This finding is consistent with other studies showing that disulfiram, in particular, helped patients to maintain abstinence and to stay in treatment for longer periods [27]. Much research has been carried out to demonstrate the efficacy of supervised low dose disulfiram as an effective treatment for alcoholism [22,28–30]. In some cases, disulfiram may be an effective and well-tolerated pharmacological treatment, within the framework of a well-integrated pharmacological, psychosocial and behavioural treatment programme [31]. Recently it has won attention as an agent adjunctive to other pharmacological medications, that specifically reduce alcohol craving [22]. Furthermore, the option of combining disulfiram with other anti-craving agents should be considered [32,33].

We know that an effective treatment of addiction is based on two important pharmacological concepts: (i) the drug used has to prove an ‘antagonist’ effect against the abused substance; and (ii) an ‘anti-craving’ effect against drug-seeking behaviour, by stimulating the substance of the abuse-related affected system [34]. These two concepts stand as the basis of methadone treatment [35]. GHB and disulfiram do not possess these two properties. GHB cannot be classified as an ‘antagonist’ drug, even if it displays some anticraving effect. Disulfiram is an ‘adversative drug’ that induces negative reinforcement, but it does not show any ‘anti-craving’ effect.

We speculate that, by combining disulfiram and GHB treatment, we can achieve a kind of ‘antagonist’ effect by using the ‘psychological threat’ of disulfiram (aversive effect), while taking advantage of the anticraving effect of GHB. In our clinical practice, we added disulfiram to GHB in the management of GHB treatment-resistant alcoholics.

The aim of the present study was to retrospectively evaluate the efficacy of a long-term (six-month) GHB-disulfiram combination treatment with reference to the ethanol intake of chronic alcoholics who met the criteria for GHB treatment resistance. We also studied endpoint clinical differences between responders and non-responder patients treated with the GHB-disulfiram combination. Lastly, we explored predictors of response to the GHB-disulfiram combination using patients‘ demographic and clinical characteristics at baseline.

To do that we compared the retention in treatment of patients treated with the GHB-disulfiram combination, for up to six months, to retention for the same subjects considering their most recent unsuccessful outpatient long-term treatment with GHB only. An additional comparison was carried out by comparing the days of complete abstention from alcohol in those two cases.

34 patients (65.4%) successfully completed the protocol and were considered to be responders; 18 (34.6%) left the programme and were considered to be non-responders. A key result was that 34 patients (65.4%) were still under treatment after six months. Table 1 shows the retention rate under treatment.

No patients left treatment without being considered non-responders. There was a major sample attrition during the first month of treatment, and a minor one during the second and the fourth. After the fourth month of treatment, no patients left the programme. Patients under long-term treatment with the GHB-disulfiram combination stayed in treatment longer (comparison: Lee-Desu statistic 13.61, df = 1, p < 0.001) than when they were treated with GHB only. More specifically, the greatest risk of leaving the GHB only therapy during participants‘ latest unsuccessful treatment occurred during the 5th and the 3th months; the next more crucial periods were, in order, the 4th, the second and the first month. No patients were in treatment after the 5th month.

Considering the days of complete abstinence from alcohol, 36 patients stayed in treatment longer with the GHB-Disulfiram combination, 12 stayed for a shorter time and four for the same time. These variations were statistically significant according to the Wilcoxon Matched-Pairs Signed-Ranks Test (z = −5.11, 2-tailed p < 0.0001).

Table 2 shows endpoint clinical differences between responder and non-responder patients. Responders differed from non-responders as to severity of illness (lower), global improvement (higher), efficacy index (more therapeutic effectiveness, no side-effects or non-interfering side-effects), GAF score (higher).

Table 3 shows predictors of response to the GHB-Disulfiram combination treatment. Sex, age, education, marital status, job, income, baseline social impairment, psychiatric comorbidity, all failed to show any effect on the retention rate. Only GHB intake (over three times a day) and a low severity at baseline predicted a better retention rate.

According to the results obtained in the present study, patients in long-term treatment with the GHB-disulfiram combination stay in treatment longer than when they are treated with GHB alone. Moreover, the GHB-disulfiram association results in a higher percentage of days of abstinence from alcohol. Responder patients compared with non-responder ones present better clinical features; are characterized by a less severe illness at baseline and an intake of GHB over three times a day.

Given the importance of dropout as a factor conditioning compliance with therapeutic programmes in the addiction field, the increased retention in treatment observed in our study stands as strong preliminary evidence in support of the GHB-disulfiram association. Moreover, the statistically significant gain in terms of days of complete abstinence from alcohol observed in patients treated with GHB-disulfiram seems to indicate a higher efficacy of this association compared with GHB alone.

As hypothesized in the introductory section, the advantages of this association can easily be attributed to the combination of the anticraving effect of GHB with the traditional adversative effect of disulfiram. However, other mechanisms have recently been taken into consideration in explaining the efficacy of disulfiram in drug dependence. These additional mechanisms, which were initially put forward to explain the efficacy of disulfiram in the treatment of cocaine addiction [39–42], have been extended to the use of this medication in treating alcohol dependence [43].

Disulfiram might be supposed to interfere with the rewarding properties of alcohol by inhibiting the noradrenergic components of craving and of highs, and be qualified as an anticraving agent. Moreover, the decrease in noradrenergic transmission mediated by disulfiram, which dampens the stress response, could also interfere with the mechanism of relapse in response to environmental influences and stressors [44–47].

The GHB-disulfiram combination might not only have a kind of ‘antagonist’ effect related to the ‘psychological threat’ or adversative effect of disulfiram and a GHB-mediated anticraving effect, but also an antirewarding effect along with a low-level, disulfiram-mediated anticraving effect. The antirewarding, disulfiram-mediated effect might be useful too against the rewarding effect of GHB (which leads to GHB abuse). This latter effect is quite similar to that described by Caputo et al. regarding the GHB-naltrexone combination [48].

An alternative interpretation is that treatment success may be interpreted not only as an effect of GHB, disulfiram, or the combination of both medications, but also of psychological modes of action, e.g. motivation, self help and self management. This study was, in fact, a controlled before/after study with participants acting as their own controls; that is, they have been compared with themselves, according to whether they were being treated with GHB alone or with a combination of GHB and disulfiram. Given the time-frame which was set to include the entire duration of the two treatments and to take into account possible differences between the sociodemographic or clinical conditions in the two periods of observation, we compared these conditions at the baseline for each of the two treatments. We found no differences. Moreover, since the design did not include a separate control group, we did not actually carry out a check on the effects of time. This meant that we were unable to differentiate between changes or effects that were measured at the end of the study, but which might have occurred anyway, without any intervention, and those that were due to the intervention itself.

The results of this study seem to indicate a greater efficacy of the GHB-disulfiram combination compared with GHB alone. Randomized controlled trials are now needed to verify this hypothesis.