A variety of medications have been studied as adjuncts to the psychosocial treatment of alcohol dependence. Given the increased interest by both government and industry, the coming decade promises to be a watershed period in the development of medications to treat alcoholism.

During this symposium, Dr. Myrick discussed the pharmacotherapy of alcoholism. Dr. Kranzler discussed the neuropharmacology of opioid antagonists and acamprosate. Dr. Anton discussed the use of opioid antagonists, both alone and in combination with other medications. Dr. Kranzler discussed depot naltrexone. Dr. Soyka discussed acamprosate studies conducted in Europe, and Dr. Chick led the final discussion.

Symposium Chairs: Henry R. Kranzler, MD, University of Connecticut School of Medicine, Farmington, CT, and Hugh Myrick, MD, Medical University of South Carolina, Charleston, SC

The FDA approval process takes many years and a minimum of $100 million to have medications come from a compound all the way through to an FDA-approved medication.

To treat substance abuse or alcoholism, there is detoxification, and then there is relapse prevention. As far as the medications for detoxification or alcohol withdrawal, the consensus guidelines and expert reports all say that the benzodiazepines are the gold standard for the treatment of alcohol withdrawal. The beta-blockers and the alpha-2-adrenergic agonists are not usually considered as monotherapy agents, and they have not been shown to block seizures that occur in alcohol withdrawal. These may be considered as adjunctive agents in the treatment of alcohol withdrawal. Using anticonvulsants as detox agents may have some advantages.

Eighty to 85% of detox occurs on an outpatient basis, and benzodiazepines are an abusable substance, so finding an agent that is not abusable would be advantageous to patients.

Lastly, some of these anticonvulsants have been shown to have not only neuroprotective effects, but neurotropic effects. This is very important in a disease process like alcohol withdrawal. Both excitotoxic damage and neural loss are occurring.

In the United States, there is limited clinical experience with the anticonvulsants. Another disadvantage is that some of the agents have hematological side effects.

Agents that have been studied for alcohol withdrawal include divalproex, carbamazepine, gabapentin, vigabatrin and levetiracetam. The most amount of information with double-blind studies is with carbamazepine, and a lot of these studies were done in Europe.

Regarding relapse prevention, advances in the pharmacotherapy of alcoholism have been enhanced by the understanding of the neurobiological substrates of alcohol. It is clear that alcohol interacts with a variety of neurotransmitter systems. Therefore, these systems may be good targets for potential therapeutic agents in the treatment of alcoholism.

Disulfiram is an aversive agent, producing an unpleasant reaction when alcohol is ingested. The mechanism behind this is the fact that it irreversibly inhibits acetaldehyde dehydrogenase, interfering with a breakdown of acetaldehyde. Because of this, there is an ethanol reaction and a variety of different symptoms, such as facial flushing, tachycardia, hypotension, nausea and vomiting, headache and physical discomfort. Some people actually die.

In many of the large double-blind studies of disulfiram, there was not a therapeutic effect compared to placebo. Only a subset of individuals may have had a therapeutic effect. Some of the more recent data have shown that there is a greater efficacy if people have supervised administration of the agent. About 15% of people can drink even though they are on disulfiram. The predictors of efficacy with disulfiram include highly motivated patients for abstinence, people with a very good social support system such as being married, behavioral contracts to take the medication and being compelled legally to take the medication.

Also of interest is targeted treatment early in abstinence, where people are given the medication when they are going into high-risk situations, such as going to a family reunion. This gives them a little extra support.

This medication is generally well tolerated. However, patients do have garlic breath, and there is some sedation from it. Peripheral neuropathy is very rare. Liver function tests need to be conducted consistently while patients are taking this medication. Get a baseline, and then conduct a test at least every 3 to 6 months thereafter.

A variety of serotonergic agents have been studied. All have different mechanisms of action, including reuptake blockade to actual receptor activity. There is good rationale for why serotonergic agents might be effective in the treatment of alcoholism. Those include the fact that serotonin is involved in a variety of things: mood, arousal, impulsivity and personality. Also, a plethora of animal research shows that these agents can block self-administration of alcohol.

These agents might work in certain subtypes of alcoholics. Type A alcoholics have a later onset, fewer childhood risk factors, less severe alcohol dependence and less psychopathology. Type Bs have earlier onset, greater psychopathology, much more impulsivity and more severe alcohol dependence. Type Bs have a lot of serotonergic dysfunction, and it seemed like they might be preferentially responsive to a serotonergic agent. However, this has not been the case. The current thinking on this is that maybe these early onset alcoholics are really super sensitive as far as the serotonin system. Using serotonergic agents almost primes these people back to drinking. A lot more research is needed in this area to really clarify when to use what agents.

Mood stabilizers can also be used. When people are detoxed from alcohol, after the 5- to 7-day period, their symptoms don’t just stop. These people complain of a variety of symptoms for weeks to months after the cessation of alcohol use. These include sleep dysregulation, irritability, mood instability and anxiety. Sleep doesn’t really normalize for up to 1 year in some of these people. If these symptoms are not addressed, they can lead to people relapsing back to use because they know how to make these symptoms better. They can just start drinking again to help their sleep and their anxiety. Mood stabilizers can treat a lot of these symptoms, and they are one of the gold standards to treat impulsivity. Additionally, many alcoholic patients have co-morbid psychiatric disorders, and these mood stabilizers can treat a variety of psychiatric disorders.

The neurochemical basis for relapse is not known, so it is difficult to know the relevant pharmacological mechanisms. However, researchers are working to identify mechanisms that are compatible or that correlate with putative mechanisms of relapse.

Naltrexone and the opioid antagonists are structurally complex and rigid, similar to the opioid receptor agonists, such as morphine. The mechanism of naltrexone is competitive antagonism at opioid receptors. Naltrexone is relatively selective for mu receptors. Other similar drugs, such as naloxone, also act as antagonists at these opiate receptors.

The pharmacology of naltrexone can explain its anti-relapse action. Endogenous opioids are released by ethanol, and these contribute to positive reinforcement so that inhibition by opioid antagonists would reduce the priming effect of alcohol. There is also some evidence that they may block the cue-induced anticipatory effects.

With acamprosate, the mechanism is much less clear. This is a flexible amino acid analog that probably has several mechanisms of action, including inhibition of glutamate release and inhibitory modulation of NMDA receptors to which glutamate binds. There are currently no drugs with a similar mechanism. Alcohol withdrawal and negative reinforcement may be a consequence of overactivity of these NMDA receptors. The acquisition and expression of conditioned responses is dependent on the glutamate system, so this is at least suggestive that this mechanism of acamprosate can explain its relapse effects.

Recent research however suggests that acamprosate may not act directly on the NMDA receptors but instead may produce effects indirectly via actions on "metabotropic" glutamate receptors. These receptors act intracellularly through second messenger systems, changing cellular excitability as a consequence. Acamprosate may interact with at least one of these metabotropic receptors, the mGluR5 receptor, which then indirectly affects the ionotropic receptor calcium flux through NMDA receptors. If true, the effects of acamprosate on the glutamate system and on alcohol withdrawal can be explained by this one mechanism.

In conclusion, naltrexone is a relatively selected mu opioid receptor antagonist. It appears to reduce alcohol reward and probably reduces relapse either due to priming effects or conditioned reward and anticipatory reward.

In contrast, acamprosate is an inhibitory modulator of the NMDA receptors perhaps indirectly via metabotropic receptors. It reduces some aspects of alcohol withdrawal, and, although not an adequate treatment for alcohol withdrawal by itself, it appears to reduce relapse due to priming and conditioned pseudowithdrawal. A simple way to differentiate the two proposed mechanisms is that naltrexone works through positive reinforcement, and acamprosate works through negative reinforcement.

Alcoholism doesn’t develop in weeks or months, but is an illness that occurs over the course of years where people continue to drink heavier and heavier. Sometimes, people are at risk, and sometimes, there are just brain changes going along with repeated alcohol consumption. The general thinking is that craving for alcohol increases over time as drinking increases. Then, after awhile, there is an accumulation of psychosocial and medical problems that leads to a full-blown alcohol-dependent syndrome. Most people level their drinking off at a certain level that persists over a long period of time before they seek treatment. When they finally decide to quit, they go through either forced or voluntary alcohol withdrawal.

During treatment, people often have episodes of relapse drinking. Episodes of relapse drinking can get bigger and bigger until there is an explosion of drinking — a binge perhaps that leads to chronic, persistent drinking. There are various triggers to relapsed drinking, such as environmental cues, changes in moods and stress, initiating craving. All are involved with the initiation of relapsed drinking.

The goal of treatment, both pharmacologic and nonpharmacologic, is to reduce the impact of these mood, stress and craving changes and environmental cues such that, over a period of time, if there is some slip drinking, it doesn’t turn into a full-blown relapse.

The brain is a complex organ, and alcohol affects almost every cell of the brain. It’s likely that a neural network or a systems approach is going to be very important in understanding alcoholism or addiction in general and how to impact it.

People who develop chronic alcoholism either because of being genetically at risk or just because of sustained and persistent heavy use ultimately wind up having brain changes. Basic scientists call them neuroadaptive changes or sensitized changes. The bottom line is that the brains of alcoholics are definitely different, both different from the brains that they had before they started drinking heavily and also different from the brains of social drinkers.

The initial stages of abstinence lead to withdrawal and its concomitant signs and symptoms. Then, later on, after a stable period of abstinence, these people have a reward memory, which is the positive aspects of drinking. This is imprinted very saliently in the brain, probably in the cortical areas. This makes these people very vulnerable for relapse. The underpinnings of these two conditions are the neurochemical systems.

Although craving is a strong intermediate between brain changes and relapse, the early cravings may be different from the late cravings.

Physicians can impact alcoholism in three major areas: the area of reward or reinforcement, the area of protracted withdrawal or the area of affective disorder or impulse control. It is fair to think about naltrexone or nalmefene and ondansetron possibly working on the reward system, where acamprosate might not only work on the reward system but also may have something to do with protracted withdrawal. The SSRIs and buspirone work on serotonergic systems, trying to stabilize affective or impulsive conditions, which may work along with the reward mechanisms to enhance relapse and persistent alcohol use.

With naltrexone, some of the earlier studies were more positive than some of the later studies. By and large, naltrexone is tolerated very well by itself and with a variety of other medications. Liver enzymes actually normalize on naltrexone rather than having a negative effect. Anything that one can do to enhance compliance in alcoholics in treatment will go a long way to improve pharmacotherapy.

The SSRIs in general have not proven so efficacious when used alone.

The COMBINE study is the only large, multisite study that’s ongoing. There are 11 sites involved. The objectives of the study are to assess the efficacy of combinations of behavioral and pharmacological interventions in the treatment of alcohol dependence. This trial is comparing naltrexone and acamprosate separately and together.

Patients who receive naltrexone are given 100 mg daily as two 50-mg tablets in the morning. Patients who receive acamprosate are given 3 g as two 500-mg tablets given three times a day. Nine cells were included in the trial. It is stratified by people getting medical management or medical management plus psychotherapy. Ultimately, there will be 1,375 people equally divided into these cells.

A pilot study was done to assess the feasibility of the protocol, to provide experience to the centers and to look at how people will tolerate the medication and if it is safe. Included in this study were 108 people using the same protocol that is going to be used in the main trial. These patients had 16 drinks per drinking day. These are moderate to moderately severe alcoholics. Most of these people were recruited from community samples, mostly through advertisements.

The medications, including the combination of naltrexone and acamprosate were well tolerated with none of the reported adverse events being significantly different from placebo. People didn’t drop out because of any of these concerns.

Overall, the retention and compliance were good, and, importantly, the combination group was no worse in this regard than either medication alone or the placebo group. Liver functions and other biological tests indicated that the medications were well tolerated. The results of this trial, which are currently under review, indicated that the main trial could proceed as planned. It is anticipated that results of the full trial will be available in about 1 to 2 years.

One of the most important things to do when one is faced with someone with alcohol dependence is to determine whether he or she has a psychiatric disorder. That may change treatment. If a patient has a bipolar disorder, there is some evidence that valproate may be useful. Some data have found that buspirone may be of some utility in these patients. For depression, one should consider the SSRIs and TCAs. Although the SSRIs have not shown a marked impact on drinking behavior, in most studies where they have been evaluated, it looks like they do treat depression as well in alcoholics as they do in nonalcoholics. Finally, if there is no psychiatric disorder, then either naltrexone or ondansetron might be used. However, ondansetron has not been widely studied, has been found to be useful mostly in early-onset alcoholics and is not approved by the FDA for the treatment of alcoholism. Some preliminary reports suggest that combining naltrexone with SSRIs or ondansetron may be of some use as a combination therapy. As always, it is important to provide some psychosocial intervention. Most of the data suggest that this is quite important both in improving compliance and in improving outcome.

Naltrexone depot is an investigational product that is not yet available for clinical use. Dr. Kranzler was principal investigator of a multi-center study to evaluate the safety and efficacy of naltrexone depot as an adjunct to motivational enhancement therapy in treatment-seeking alcoholics. The design was an initial 12-week double-blind randomized placebo-controlled trial with 315 alcohol-dependent subjects. Compliant subjects who received injections during the first 3 months and were at least 75% compliant with the data collection visits were eligible to continue for an additional 9 months, for a total of 12 months.

The initial screening period required that subjects have 3 consecutive days of self-reported abstinence. Those who were able to achieve that were then given oral naltrexone 50 mg a day for 4 days to identify naltrexone intolerance. Abstinence was not required during this oral naltrexone period. Patients were then randomly assigned to receive injections of either naltrexone or placebo, which were administered under double-blind conditions. The visit schedule made it possible to monitor adverse effects and to quantify drinking behavior using the timeline follow-back method. Following randomization, there were eight visits during the first 12 weeks — weekly for 4 weeks and biweekly for 8 weeks. Then, during the extension period, there were monthly visits.

The study medication is a microsphere polylactide formulation that is biodegradable. In the active condition, it was impregnated with naltrexone, and in the placebo condition, it consisted only of the polylactide microspheres.

The first injection was a 300-mg dose, so 150 mg was injected in each buttock. Each subsequent monthly injection was 150 mg administered as a single injection. The psychosocial treatment was motivational enhancement therapy (MET) — five 1-hour sessions during the first 12 weeks. Patients were also encouraged to attend self-help group meetings. The analysis was an intent-to-treat analysis, with missing data or data from subjects who discontinued the study treated conservatively as heavy drinking days.

Analysis of variance was used to compare groups on continuous variables. Survival analysis was used for time to event measures. Fisher’s exact test was used for categorical variables.

The sample was roughly evenly distributed between the two conditions. The age of onset of alcohol problems was in patients’ early 30s. They had not had many treatments — the range was zero to four prior alcohol treatments, and the mean was less than one. During the pretreatment period, subjects had approximately 2 of 3 heavy drinking days, or 20 out of the preceding 30 days.

There were no major safety differences between the two conditions after 3 months of treatment. There were seven serious adverse events in the placebo group — four of these were alcohol-related, one was related to depression, one was related to surgery (the FDA requires that hospitalization for any reason be considered a serious adverse event) and one involved cellulitis at the injection site. In the naltrexone depot group, there were two serious adverse events — one was a fracture and one was a human bite.

In both groups, more than 80% of patients had one or more adverse events. Upper abdominal pain was significantly more common in the naltrexone group, which is consistent with what is known about the pharmacology of the opiate antagonists and the distribution of opiate receptors. Interestingly, irritability was more common in the placebo group.

In terms of injection site reactions, a greater percentage was seen in the naltrexone group. The difference was primarily on induration, which was significantly more common in the naltrexone group. There was a trend for inflammation to be more common in the placebo group, although it did not reach significance.

Regarding nonheavy drinking, there was very little separation between groups. At the end of the 3-month trial, they were not different at all. There was a trend for a difference at 6 months, but it did not reach significance. In contrast, regarding cumulative days of abstinence, there was a trend at 3 months, which reached significance at 6 months.

In terms of time to event, the naltrexone group had a greater time to relapse to any drinking and to heavy drinking, effects that were seen during the initial 3 months and that persisted throughout the 6 months of treatment. Regarding the percentage of subjects abstinent through 3 months, there was an effect for all subjects, with the naltrexone group having a better outcome. This effect was also evident through 6 months, with roughly a nine-fold increase in the likelihood of abstinence during this period. Patients who were abstinent at randomization were more likely to stay abstinent if they got naltrexone depot than if they got placebo. This suggests that patients who are abstinent on oral naltrexone may be a particularly good group in which to use a depot naltrexone formulation for maintenance therapy.

At 6 months, the percentage of subjects with no heavy drinking also favored naltrexone.

In conclusion, the naltrexone depot was shown to be safe and well tolerated compared to placebo. Although it did not significantly reduce cumulative heavy drinking days, it did appear to increase cumulative days of abstinence, and it increased the time to relapse for both drinking and heavy drinking. It also increased the likelihood of total abstinence.

The studies that have been done in Europe are very different from the ones done in the United States. In Europe, there is a different relapse definition. Relapse means relapse to drinking, not to heavy drinking. The psychotherapy of alcoholism in Europe is pretty much abstinence-based and abstinence-oriented.

In Western Europe, there are approximately 10 million alcohol-dependent patients. Five to seven percent of the population is alcohol dependent. A low number of people do not drink alcohol at all. In Scotland, 93% of people drink some sort of alcohol. It’s the same in Germany. This is certainly different from what we know in the United States.

Approximately a dozen clinical placebo-controlled randomized trials have been conducted in Europe. Eleven trials are reasonably comparable for criteria and demographic variables like clinical history of patients and the major efficacy outcome criteria. Safety data are relatively comparable for nine studies. The treatment duration in the trials varied. A few had a 3-month treatment phase, while most of them had a 6-month treatment phase. Five trials had a treatment phase of more than 1 year.

A few trials had a medication-free follow-up period, which varied from 4 weeks to 12 months. The assessment intervals during treatment differed in the European trials — days 0, 30, 90, 180, 270 and 360 in the longer trials. All these studies were placebo-controlled. They were performed basically under naturalistic conditions, and there were, with few exceptions, two treatment groups — acamprosate and placebo in nine trials.

All patients started on medication after the period of acute detoxification, usually 1 to 4 weeks after last alcohol intake. All patients were abstinent at the start of the treatment. Studies were carried out in accordance with European good clinical practice standards. Central laboratory facilities were used. The primary outcome criterion was time to first relapse.

Approximately two-thirds of the patients in all of the studies were male, and the mean age was about 40 to 45 years old in most studies.

Taken together, there was an abstinence rate in the acamprosate groups of 35% at the end of the treatment versus 21% in the placebo groups. All but one study showed significant results in favor of acamprosate, but there are huge differences between studies in relation to abstinence rates and in relation to patients missing.

Regarding time to first drink, in most studies, there was a clear and significant result in favor of acamprosate. Regarding the percentage of abstainers in patients on treatment in the five trials that have a 1-year treatment phase, there was a more than 10% difference between the acamprosate group and the placebo group at the end of the trial.

Regarding the proportion of days abstinent during treatment, the acamprosate groups were superior to the placebo groups in all but one study. When comparing the percentage of patients who were abstinent, again, there was a significant result in favor of acamprosate.

However, acamprosate causes quite a few side effects. The most frequent side effect is diarrhea, which vanishes in most patients after a few days. Very few patients have dropped out of treatment because of adverse events. Gastrointestinal irritation was 15% in the first 30 days and 8% in days 31 to 90. It is 4% to 8% higher than in the placebo groups. Sex drive both increased and decreased with acamprosate. There is no clear explanation for this. Nine percent experienced this in the first 30 days of treatment in the acamprosate groups, 3% more than the placebo groups. Difficulty falling asleep was reported in 7% of patients between days 180 and 270, which was 3% more than in the placebo group.

Taken together, there is a lot of evidence for a comparatively good safety profile. There is no risk for alcohol interactions with acamprosate. Apparently, there is no risk for an overdose.

In conclusion, the abstinence rate is nearly doubled in the acamprosate groups compared to placebo. Using a psychosocial support program, acamprosate significantly improved the abstinence rate following alcohol withdrawal. Acamprosate is equally effective for the range of psychosocial support strategies. In the meantime, following these placebo-controlled trials, a number of phase 4 studies have been conducted in Europe that basically confirm the abstinence rates found in the placebo-controlled trials.

The U.K. Health Technology Board is an agency funded by the government that examines technologies and advises the national health service on what is cost effective. Alcoholism is a serious problem in Scotland, and Scotland has given it a high priority.

In a meta-analysis of published and unpublished studies that included 4,500 patients in randomized, controlled studies of acamprosate, some negative outcomes were found. The number of patients you need to give a treatment to in order to have one patient make a good recovery is 13.3. The naltrexone literature shows the number needed to treat is 10.4.

With statins, there is a little confusion about the number needed to treat to prevent one myocardial ischemic event, but it is somewhere between 60 and 100.

How do these compare with psychosocial treatments? The results of the meta-analysis have been extrapolated to 1,000 treated patients. In Scotland, out of 1,000 patients, only 150 have gotten a good result. That includes patients who have done some drinking but haven’t had a serious relapse to heavy drinking. Fifteen percent is the expected baseline rate with baseline treatments.

Using acamprosate, the success rate can be expected to increase by another 80 patients. Unsupervised disulfiram does not produce a much greater effect than basic support. Supervised disulfiram looks a bit better.

Researchers in Scotland have examined the cost of standard treatments. There would be a very impressive cost savings to Scotland if it had the trained personnel to give psychosocial treatments, such as coping skills methods and family-marital behavior therapy.

The cost of acamprosate treatment turns out to be cost neutral. This is because, in Scotland, doctors are needed to prescribe this. Psychiatrists are used in the usual methodology of acamprosate prescribing, and patients are also traditionally having some psychosocial treatment. There is no added cost to the health service of offering acamprosate treatment, and this says nothing about the benefits to families and the individual.