Naltrexone

Naltrexone is an opioid receptor antagonist. Alcohol is thought to be reinforcing because it stimulates release of endorphin (an endogenous opioid) in the brain, which leads to activation of the dopaminergic “pleasure-reward-motivation” pathway (see Chapter 2). Naltrexone blocks this activation, which is hypothesized to explain its clinical effectiveness. In addition, other mechanisms, such as reducing impulsivity and modulation of the HPA axis and stress system, have also been proposed to be involved in mediating naltrexone’s effects.

First approved for the treatment of alcohol dependence in 1994 by the U.S. Food and Drug Administration, naltrexone has been subject to numerous randomized and controlled clinical trials in which small to medium effect sizes have been reported (Jonas et al., 2014; Maisel et al., 2013; Rösner, Leucht, Lehert, & Soyka, 2008). It is now licensed in many countries in Europe, including the UK, as well as in Australia. At the dose commonly used, 50 mg daily, naltrexone has been shown to reduce relapse rates in alcohol dependent patients in combination with psychosocial treatment (for references, see Donoghue et al., 2015; Jonas et al., 2014; Lingford-Hughes et al., 2012; Maisel et al., 2013; National Institute for Health and Clinical Excellence, 2011). For this effect, the number needed to treat (NNT) has been calculated at about 12 (Jonas et al., 2014). However, naltrexone does not appear as effective in maintaining abstinence, with an NNT of 20. Therefore, naltrexone may be better used in those patients who have the occasional drink or lapse to prevent a full-blown relapse rather than to sustain abstinence (Maisel et al., 2013). Naltrexone has also been shown to reduce the euphoria or pleasure associated with alcohol intake and, in some studies, to reduce craving. No single psychosocial approach has shown a particular advantage over another when combined with naltrexone.

In relapse prevention trials, naltrexone is started once someone has stopped drinking. Treatment is usually for 3–6 months in the first instance, but should be reassessed as to whether to continue if the patient has relapsed. Although earlier studies suggested that the beneficial effects of naltrexone may not persist, later studies such as COMBINE reported continued benefit persisting for up to a year (Donovan et al., 2008). Although not everyone benefits from naltrexone, no robust predictors have been identified to guide clinical decision-making, although some studies have found potential indicators worthy of replication studies (Bogenschutz, Scott Tonigan, & Pettinati, 2009). Although a functional polymorphism, Asp40 allele, of the mu opioid receptor gene had been shown to predict naltrexone treatment response in a range of studies (e.g., COMBINE; Anton et al., 2006), a more recent prospective trial did not find such an effect (Oslin et al., 2015).

Naltrexone is generally a well-tolerated and safe medication. Side effects include nausea, vomiting and abdominal pain, headache, reduced energy, joint and muscle pain, and sleeping difficulty. Loss of appetite, diarrhea, constipation, increased thirst, chest pain, increased sweating, increased energy, irritability, chills, delayed ejaculation, and decreased potency are less frequent side effects. Although concerns about hepatic toxicity have been raised, this occurred with much higher doses (>300 mg/d) than the dose generally used to treat alcohol dependence (50 mg/d). Indeed, liver function test improvement is often seen in trials of naltrexone as alcohol consumption reduces. Nevertheless, naltrexone should not be used in those with acute liver failure and should be used cautiously when serum aminotransferases are four to five times above normal. It should not be prescribed to women who are pregnant or breast-feeding.

Most people tolerate the daily 50 mg dose, although it may be sensible to take a lower dose (25 mg) for the first 3–4 days of treatment to minimize side effects, particularly nausea in women. More recent trials of naltrexone have used 100 mg/d rather than 50 mg/d, but this higher dose has not been shown to be clearly more effective and may incur more side effects (Jonas et al., 2014). Naltrexone can be safely taken with many other medications but, as an opioid antagonist, it cannot be taken by those requiring opioid analgesia. Careful assessment of whether the patient is taking prescribed or over-the-counter opioid medications is therefore essential. If in an emergency analgesia is required, nonopioid medication will have to be used because blockade due to naltrexone may last up to 72 hours. Patients should be encouraged to carry a card stating that they are taking naltrexone in case of such an emergency.

Patients do not have to be abstinent before they commence naltrexone, but outcomes are better when they are (Maisel et al., 2013). Unlike for disulfiram, it is safe to drink with naltrexone. Indeed, it has been suggested that due to naltrexone’s hypothesized mechanism of action in reducing the pleasurable effects of alcohol, naltrexone may be beneficial in reducing consumption in those who are still drinking. The few studies of naltrexone in those who are less dependent or who are “heavy drinkers” have shown some benefit, including recently in younger drinkers (e.g., Heinälä et al., 2001; Kranzler et al., 2009; O’Malley et al., 2015), although the evidence is insufficient to justify widespread clinical use.

Naltrexone has also recently become available in a long-term injected formulation. Due to the once-a-month dosing, this approach may improve treatment compliance, thus helping the patient whose motivation to stop drinking waxes and wanes frequently. An early trial was positive and was also associated with improvements in quality of life, but a large amount of clinical experience has not yet accumulated (Garbutt et al., 2005; Pettinati et al., 2009). Because injectable naltrexone has shown impressive benefits when used with heroin dependent patients, its use should be considered for any problem-drinking patients who are also dependent on opioids (Krupitsky et al., 2012).

Nalmefene

Nalmefene is also an opioid antagonist whose first trial as a treatment for alcoholism took place in the 1990s. But it has only recently been licensed for this indication in several countries in Europe, including the UK. Whereas naltrexone is an antagonist at mu and kappa receptors, nalmefene is an antagonist at the mu opiate receptor and a partial agonist at the kappa receptor. The impact of this on the clinical efficacy of nalmefene is not clear, but preclinical data show a greater role for kappa compared with mu receptors in reducing consumption in dependent animals (Walker, Zorrilla, & Koob, 2011). No clinical trials have compared nalmefene with naltrexone.

In two European trials, nalmefene reduced heavy drinking days and total alcohol consumption in alcohol dependent patients at 6 months (Gual et al., 2013; Mann, Bladström, et al., 2013). The psychosocial intervention used combined motivational and adherence-enhancing elements to support behaviour change. Alcohol dependent patients did not have to be sober before entering the trial and also were instructed to take nalmefene “as-needed” (i.e., when they perceived there was a risk of drinking). Most patients took it at least 4 days per week. Further analyses of these trials showed that the effect of nalmefene was not evident in those who had reduced their drinking levels between screening and randomization. Thus, nalmefene benefited those who were still drinking at a high risk level (>60 g/d for a man; >40 g/d for a woman) (van den Brink et al., 2013). Nalmefene is therefore licensed to reduce alcohol consumption in alcohol dependent patients “who continue to have a high drinking risk level two weeks after initial assessment, without physical withdrawal symptoms and who do not require immediate detoxification.”

Greater adverse effects such as nausea, dizziness, insomnia, and fatigue were also reported for nalmefene than placebo. These side effects were generally mild to moderate and present only for first 1–2 days, but nonetheless contributed to treatment dropout.

Acamprosate

Acamprosate (calcium bis-acetyl homotaurinate), a simple derivative of the amino acid taurine, has been used in the treatment of alcohol dependence in Europe for more than 20 years and more recently became available in the United States. Its target in the brain is uncertain; however, it does modulate glutamatergic N-methyl-D-aspartic acid (NMDA) receptor transmission, probably involving metabotropic glutamate receptor subtype 5 (mGluR5) antagonism. Acamprosate may have indirect effects on gamma-aminobutyric acid (GABA_A) receptor activity as well (Kalk & Lingford-Hughes, 2014). It is thought to modulate alcohol withdrawal effects and limit any aversiveness associated with the cessation of drinking.

Numerous clinical trials of acamprosate in combination with a psychosocial intervention have been conducted throughout the world, with a particular concentration in Europe. Meta-analyses have shown acamprosate significantly improves abstinent rates, with a NNT of around 12 to prevent return to any drinking (Donoghue et al., 2015; Jonas et al., 2014; National Institute for Health and Clinical Excellence, 2011; Rosner et al., 2010). Acamprosate’s ability to reduce quantity and frequency of drinking is less certain. Therefore, acamprosate’s main effect is to maintain abstinence rather than prevent a lapse becoming a relapse, which is thus the opposite of naltrexone’s (Maisel et al., 2013). Acamprosate tablets (333 mg each) are taken in doses of either 1,998 mg or 1,332 mg daily, depending on the patient’s weight. The number and size of pills may present challenges with adherence for some patients.

Like naltrexone, in clinical trials, acamprosate is started once the person is abstinent. U.S. (COMBINE) and UK trials have suggested that poorer outcomes are seen with longer periods of abstinence before starting acamprosate (Chick, Howlett, Morgan, & Ritson, 2000; Gueorguieva et al., 2015). Given the potential neuroprotective effect of acamprosate and that it can be given safely along with medication for detoxification (see Chapter 11), acamprosate could be started before or during detoxification rather than after (Lingford-Hughes et al., 2012). Treatment with acamprosate is generally recommended for up to 6 months, although license suggests 1 year, and its beneficial effects may persist after stopping (Rosner et al., 2010). If abstinence is not maintained and drinking behaviour is not changing after 4–6 weeks, it is wise to consider whether acamprosate should be continued (National Institute for Health and Clinical Excellence, 2011). Although earlier reports suggested that those with anxiety or more severe dependence were more likely to respond to acamprosate, this was not found in reviews (e.g., National Institute for Health and Clinical Excellence, 2011; Verheul et al., 2005).

Acamprosate is generally a well-tolerated and safe medication. Its main side effects are related to the gastrointestinal system such as nausea and diarrhoea. Acamprosate should not be prescribed to individuals with significant renal impairment or hepatic failure, nor to women who are pregnant or breast-feeding. It does not interact with alcohol or diazepam, appears to have no addictive potential itself, and can be used safely with a wide range of medications, including antidepressants.

Naltrexone versus Acamprosate

Most clinical studies compare active medication with placebo, but some have directly compared acamprosate and naltrexone. A European study found that naltrexone (50 mg/d) was superior to acamprosate (1,998 mg/d) or placebo (Kiefer et al., 2003). but an Australian study using the same doses found that neither were superior to placebo (Morley et al., 2006). Combined naltrexone and acamprosate has been shown to improve drinking outcomes compared with acamprosate alone but not to naltrexone alone (Kiefer et al., 2003). One of the largest studies ever conducted in alcohol dependence, COMBINE, a nine-arm U.S. study, compared acamprosate (3 g; i.e., a larger dose than is now licensed) with naltrexone (100 mg) individually or together in addition to standard “medical management” or more intensive combined behavioural intervention (CBI) (Anton et al., 2006). Naltrexone with medical management alone or in combination with CBI resulted in greater improvements than placebo or medical management alone, whereas acamprosate showed no evidence of additional efficacy in any combination. Combined acamprosate and naltrexone conferred no benefit. Modeled on COMBINE, PREDICT was a German study that found neither medication was superior to placebo (Mann, Lemenager, et al., 2013). Meta-analysis of all these studies found no difference in global effectiveness between acamprosate and naltrexone (Jonas et al., 2014). However, another meta-analysis of clinical trials reported that effect sizes for naltrexone tend to be larger than those for acamprosate when the outcome is reduced heavy drinking and craving, but acamprosate has larger effect sizes when the measured outcome is abstinence (Maisel et al., 2013).

What contributes to these disparate findings of whether naltrexone or acamprosate are effective, and does it inform clinical decision-making? There are more trials of naltrexone conducted in the United States than elsewhere, whereas most trials of acamprosate are conducted in Europe. Indeed the few U.S.-based studies did not find acamprosate efficacious, whereas European trials did find naltrexone efficacious. There has been extensive discussion of these different results from U.S. and European trials (Donoghue et al., 2015). In European trials, participants are typically recruited from in-patient programmes and treatment services, whereas U.S.-based trials usually advertise for participants who have to be able to be sober for 3 days. U.S. participants are therefore less likely to be severely dependent, and this, along with the differences in healthcare settings, is likely to contribute the variability in outcomes seen.

Baclofen

Baclofen is a GABA_B receptor agonist licensed to treat muscle spasms in disorders such as multiple sclerosis. Evidence is accruing to support a role for baclofen in treating alcohol dependence, with trials in Italy and Germany reporting reduced craving, anxiety, and relapse rates (Addolorato et al., 2007; Müller et al., 2015). A U.S. study failed to find superiority of baclofen over placebo, but, as discussed earlier, this was likely due to differences in participants (Garbutt et al., 2010). In the U.S. study, patients were less severely alcohol dependent and the majority did not desire abstinence, unlike in the two European trials.

Baclofen is started after detoxification and is titrated up from 5 mg three times a day. The optimal dose of baclofen is unknown: some alcohol dependent participants require 10 times the dose used in the first trial (i.e., 300 mg/d vs. 30 mg/d). Although some appear to tolerate such large doses, side effects such as sedation and withdrawal seizures become more likely over time. Larger doses have been investigated in recent studies, with one allowing up to 270 mg, although only a third of patients reached this amount (Müller et al., 2015). Other studies are due to report shortly to help inform clinicians about the optimal dose of baclofen, as well as its use in other substance dependencies such as nicotine or cocaine alone and when comorbid with alcoholism.

Topiramate

Topiramate is an anticonvulsant affecting many systems in the brain including glutamate, GABA, and dopamine. A meta-analysis of trials reported that topiramate reduced heavy drinking and increased abstinence (Blodgett, Del Re, Maisel, & Finney, 2014). Notably, in some of these trials, topiramate was started while the patient was still drinking rather than after detoxification.

Side effects reported included paresthesia/numbness, nausea/vomiting, and cognitive impairment, particularly at higher doses and without slow enough titration. More research on dosing regimes is clearly needed. Topiramate is also being investigated as a treatment for stimulant use disorders.

Pregabalin and gabapentin

Despite their names, pregabalin (Lyrica) and gabapentin (Neurontin) do not directly target the GABA system but instead are calcium channel modulators. They are licensed for treating a range of disorders such as seizures, neuropathic pain, and generalized anxiety disorder (pregabalin only). These medications are being evaluated as possible treatments for alcohol and other substance dependence. Both medications have been studied for treating alcohol withdrawal, but the evidence is not promising (Guglielmo, Martinotti, Clerici, & Janiri, 2012; Leung et al., 2015). However, some studies suggest that gabapentin and pregabalin could aid in relapse prevention, although the evidence is insufficient to recommend them for routine use over other medications (Pani, Trogu, Pacini, & Maremanni, 2014). There are concerns about the abuse liability of these medications and also their safety in combination with other central depressants such as alcohol and opioids. Trials thus far have reported, however, that these medications are well-tolerated and have a favourable safety profile.

Depression

Three meta-analyses of trials of antidepressants in treating comorbidity all suggest that antidepressants may reduce depressive symptoms but not necessarily alcohol consumption in depressed alcohol dependent patients (Iovieno et al., 2011; Nunes & Levin, 2004; Torrens, Fonseca, Mateu, & Farré, 2005). Mixed serotonin-noradrenergic antidepressants appear superior to serotonin reuptake inhibitors (SSRIs). There are no placebo-controlled trials of the newer antidepressants. Meta-analytic reviews indicate that antidepressants are less effective in treating depression when the patient is still drinking and that stopping or considerably reducing drinking generally leads to a substantial improvement in mood. It is therefore better to assess the need for an antidepressant once abstinence (or, failing that, significantly reduced consumption) is achieved.

Concerning relapse prevention medication, in one of the largest trials in comorbidity, disulfiram and naltrexone alone and in combination were compared with placebo in patients with a psychiatric comorbidity that included depressive, anxiety, and psychotic disorders (Petrakis et al., 2005). A secondary analysis comparing those with and without depression concluded that naltrexone and disulfiram can be used safely in depressed alcohol dependent patients with comparable efficacy to those not depressed (Petrakis et al., 2007). A recent study comparing naltrexone (100 mg) or sertraline (200 mg) alone or in combination found that only the combination improved drinking outcomes and mood (Pettinati et al., 2010).

The reviews and trials support the commonly given clinical advice to not start an antidepressant in a patient while they continue to drink heavily and to assess their depression and need for antidepressant once abstinent, ideally for at least 2–3 weeks (Lingford-Hughes et al., 2012; National Institute for Health and Clinical Excellence, 2011). There is also support for using relapse prevention medication, with the most evidence available for naltrexone, although an adverse impact on outcome has not been seen for any such medication.

Anxiety disorder

Careful assessment is required to establish whether a comorbid anxiety disorder is present because, if left untreated, reducing alcohol consumption is less likely. As with depression, because treatments have limited impact on the anxiety disorder with continued drinking, abstinence should be encouraged (Lingford-Hughes et al., 2012; National Institute for Health and Clinical Excellence, 2011). In post-traumatic stress disorder (PTSD), naltrexone or disulfiram alone or in combination have been shown to result in greater improvements in alcohol consumption than placebo (Petrakis et al., 2006b). A later study of combat veterans with PTSD, however, did not show any additional benefit of naltrexone when combined with either desipramine or paroxetine (Petrakis et al., 2012).

Bipolar disorder

There is limited evidence to suggest the optimal mood stabilizer regimen for patients with drinking problems and bipolar disorder (Lingford-Hughes et al., 2012). Concerning relapse prevention medication, preliminary reports suggest no particular benefit with acamprosate although also no worsening in mental state. A secondary analysis of Petrakis’s study of naltrexone and disulfiram compared those with a psychotic disorder, of which the majority (73 percent) had bipolar disorder, to participants without a psychotic disorder and reported that medication was superior to placebo (Petrakis, Nich, & Ralevski, 2006a). Given the potential of disulfiram to increase psychotic symptoms, naltrexone is easier and safer to use in bipolar disorder.

Schizophrenia

The few studies of relapse prevention medication in schizophrenia identified no adverse impact on mental state but no particular benefit to drinking outcomes either. Although disulfiram has been safely used in schizophrenia, it is however best used in a specialist setting once other options have been exhausted. The limited evidence available suggests there is no benefit from the second- over the first-generation antipsychotics (Lingford-Hughes et al., 2012; Petrakis et al., 2006a). Although there are no prospective trials, some evidence suggests that clozapine improves drinking outcomes. Given its beneficial effect in improving treatment-resistant schizophrenia, clozapine should be considered in those patients with problem drinking and whose schizophrenia persists despite adequate antipsychotic medication.