The COMBINE study was only one trial designed by academics to maximize internal scientific validity. It excluded individuals with other significant psychiatric and medical illnesses (more often the rule than the exception in some clinical settings)-individuals deemed too severely ill or who needed hospitalization.
There are some important questions that clinicians and third-party payers always ask: Do medications to treat alcoholism really work-and if they do, can they be used alone or with behavioral therapies or other interventions? Do medications work any better than behavioral interventions? These are very practical questions . . . ones that experienced clinical researchers tried to address a few years ago in the NIH’s National Institute on Alcohol Abuse and Alcoholism (NIAAA)-sponsored multicenter clinical trail, the COMBINE (Combining Medications and Behavioral Interventions for Alcoholism) study.1
Project MATCH (Matching Alcoholism Treatments to Client Heterogeneity), another NIAAA-sponsored trial, showed that several different types of psychosocial interventions, such as cognitive-behavioral therapy (CBT), motivational enhancement therapy (MET), and a 12-step facilitation approach, appeared to be equally effective in promoting abstinence and reducing relapse when delivered by qualified and trained counselors.2 In addition to interventions used in Project MATCH, several medications have been shown to be effective in the treatment of alcohol dependence and were approved by regulatory authorities in the United States and Europe.
Naltrexone, an opiate antagonist, had shown efficacy in a number of studies when combined with psychosocial interventions that included CBT, and there was some evidence that the drug worked best in this context.3 In addition, findings from European trials suggest that acamprosate, a brain glutamate stabilizer, helps maintain abstinence post-detoxification across a range of ancillary treatments that had not been well defined.4 In the United States, a multicenter trial did not show acamprosate to be more efficacious than placebo in the context of a medical model supportive/educational approach.5
Given this background, the COMBINE study was designed to address several important questions:
• Do naltrexone and acamprosate produce effects that are different from those of placebo in a well-defined outpatient group of treatment-seeking alcohol-dependent individuals?
• Since naltrexone and acamprosate have different, and perhaps complementary, modes of action, do they work better when combined or when either one is used alone?
• Is the combination of moderately intensive behavioral intervention/counseling and either naltrexone or acamprosate more efficacious than using these agents with a medical management approach (discussed in greater detail below)?
• How do the outcomes of a group that receives behavioral intervention alone compare with those of a group that receives either active drug or placebo?
The COMBINE study recruited 1383 persons with alcohol dependence, mostly derived from community volunteers not currently in treatment, across 11 academic centers. The initial assessment lasted 4 to 6 hours over several days, and the volunteers were required to remain abstinent for a minimum of 4 days. This was required because previous studies done with both naltrexone and acamprosate were initiated after a period of abstinence. From a clinical point of view, patients may not tolerate these medications while drinking; withdrawal symptoms may also complicate their use. The volunteers were randomized to receive naltrexone (up to 100 mg daily), acamprosate (3 g daily), a combination of the two drugs, or placebo for 16 weeks. All participants also received medical management with or without combined behavioral intervention (CBI).
Medical management was developed to provide an approach that could be used in an addiction treatment setting or primary care setting and could be delivered by various health care providers (nurses, physicians, physician assistants, etc). Medical management-similar to that provided in the treatment of diabetes, hypertension, etc-involved up to 9 sessions of medical education, advice, compliance enhancement, feedback on laboratory results, and adverse-effect review, as well as Alcoholics Anonymous attendance encouragement.6,7
CBI consisted of a patient-centered approach in which aspects of MET, CBT, and 12-step facilitation were delivered by trained and supervised counselors and that was based on individual needs but drawn from a limited procedure manual.8 To evaluate whether drinking outcomes were better for the groups receiving the active drugs or placebo and/or medical management, one group was randomized to CBI only, without any drugs and with limited medical management.
The results were in some ways unexpected but important.
The main positive finding was that individuals treated with naltrexone did significantly better than the placebo group on various drinking measures-but only when they received medical management alone. When they received CBI in addition to medical management, the difference between the placebo and naltrexone groups was no longer evident.
These results are best illustrated by a good composite outcome measure, defined as abstinence or moderate drinking without alcohol-related problems during the last 8 weeks of the trial (Figure 1). The number of individuals with alcohol dependence that need to be treated (number needed to treat [NNT]) to get 1 response that was more favorable with naltrexone than with placebo was 6. This compares favorably with the NNT in studies of chronic depression, Crohn disease, type 2 diabetes mellitus, and Alzheimer disease.
Important negative findings indicated that:
• Acamprosate was no better than placebo-with or without CBI
• Acamprosate added to naltrexone was not significantly better than naltrexone alone
• CBI was less effective when given without active drug or placebo or medical management.
After 1 year of follow-up, the effects observed during the treatment period were still observable but for the most part were no longer significant. This suggests that at least some individuals relapsed to heavier drinking once treatment had stopped.
A cost-effectiveness analysis showed that the best combination of costs and effectiveness was observed in the naltrexone-treated group, which also received medical management.9
Three years later, the treatment provided during the study remained effective in reducing social costs (health care utilization, arrests, and motor vehicle accidents) in these alcohol-dependent individuals.10
The COMBINE study investigators had planned to evaluate the effects of genes on treatment response and collected DNA samples from willing participants. During the trial, it had been reported by researchers not involved in the study that a single nucleotide polymorphism (SNP) in the µ-opioid receptor (OPRM1) gene might be associated with better naltrexone response.11 In a post hoc analysis of the COMBINE study, this asparagine to aspartate (Asn40Asp) OPRM1 SNP, found in 22% of the white participants, was associated with the best response to naltrexone.12 Using the Clinical Global Outcome measure, one can see this effect in Figure 2.
The COMBINE study was important for what it discovered and for the questions it raised. First of all, it was surprising that the study did not find acamprosate efficacious, considering the many positive European studies.4 The lack of effect, however, was consistent with another multicenter study performed in the United States.5 It is possible that alcohol-dependent individuals who for the most part had not experienced alcohol withdrawal symptoms or who were not sufficiently abstinent during acamprosate induction might not have been the best candidates for this medication. Also, the fact that naltrexone did not add anything to CBI therapy (as long as participants were receiving either active drug or placebo and seeing a health care provider) was a bit puzzling, given that naltrexone has been used to augment CBT.3
CBI may have been more effective than routine CBT and therefore may have provided maximal treatment effectiveness so that naltrexone did not add additional effectiveness. The fact that CBI was not as effective when individuals were not seeing a health care provider and taking either active drug or placebo is worth noting but must be interpreted in light of the fact that participants, on balance, expected to receive “pills” when they volunteered. Whether this effect was a negative expectancy effect or whether seeing a health care provider in addition to a therapist is important for maximal treatment effectiveness cannot be elucidated more clearly from the study design.13
What is the take-home message for clinicians? The COMBINE study and many other clinical trials have shown that there are many people living and working in communities across the United States who are willing to volunteer for alcohol-dependence treatment trials. Many of these individuals have never before received treatment and many have not been asked by, or have not been honest with, their primary care providers and/or psychiatrists about the extent of their alcohol use. Specifically, however, the COMBINE study showed that health care providers can either treat alcohol-dependent individuals with naltrexone and medical management or refer them to a trained counselor who might use the types of techniques prescribed in the CBI manual.7
There are several caveats. Medical management consisted of seeing the patient 9 times over 4 months (initially once a week and then every 2 to 4 weeks). Each 15- to 20-minute session emphasized medication adherence and provided support, education, and laboratory monitoring (eg, feedback about biomarkers of drinking). There is no evidence that providing medical management less frequently, or for less time, would be as effective. However, this might be individualized depending on the severity of alcohol addiction, the degree of medication adherence, and the patient’s response.
Those who received CBI alone (or by extension, other moderately intensive weekly counseling sesions) did not fare as well. While we cannot say for sure whether just seeing a health care provider without taking a pill would be enough to enhance the counseling treatment, the suspicion would be that the more skilled support provided to the patient, the better the expected outcome.
While the evidence for the differential response to naltrexone based on genotype is very promising and important, it is still too early to prescribe this approach given the relatively limited post hoc information that is available. Prospective studies currently under way should provide more definitive information.
The COMBINE study was only one trial designed by academics to maximize internal scientific validity; results cannot be extrapolated to all populations and individuals. For instance, it excluded individuals with other significant psychiatric and medical illnesses (more often the rule than the exception in some clinical settings)-individuals deemed too severely ill or who needed hospitalization. In addition, not all communities have trained addiction professionals or medication prescribers to deliver the types of treatments offered in the COMBINE study.
The options for treating individuals with alcohol use disorders is expanding. We can hope that advances in addiction neuroscience will pave the way for new and improved treatments.
1. Anton RF, O’Malley SS, Ciraulo DA, et al; COMBINE Study Research Group. Combined pharmacotherapies and behavioral interventions for alcohol dependence: the COMBINE study: a randomized controlled trial. JAMA. 2006;295:2003-2017.
2. Matching alcoholism treatments to client heterogeneity: Project MATCH posttreatment drinking outcomes. J Stud Alcohol. 1997;58:7-29.
3. Berglund M, Thelander S, Salaspuro M, et al. Treatment of alcohol abuse: an evidence-based review. Alcohol Clin Exp Res. 2003;27:1645-1656.
4. Mann K, Lehert P, Morgan MY. The efficacy of acamprosate in the maintenance of abstinence in alcohol-dependent individuals: results of a meta-analysis. Alcohol Clin Exp Res. 2004;28:51-63.
5. Mason BJ, Goodman AM, Chabac S, Lehert P. Effect of oral acamprosate on abstinence in patients with alcohol dependence in a double-blind, placebo-controlled trial: the role of patient motivation. J Psychiatr Res. 2006;40:383-393.
6. Pettinati HM, Weiss RD, Dundon W, et al. A structured approach to medical management: a psychosocial intervention to support pharmacotherapy in the treatment of alcohol dependence. J Stud Alcohol Suppl. 2005;15:170-178.
7. National Institute on Alcohol Abuse and Alcoholism. COMBINE Monograph Series. http://pubs.niaaa.nih.gov/publications/COMBINE.htm. Accessed March 16, 2011.
8. Miller WR, ed. Combined Behavioral Intervention Manual: A Clinical Research Guide for Therapists Treating People With Alcohol Abuse and Dependence. Vol 1. Bethesda, MD: US Dept of Health and Human Services; 2004. NIH publication 04-5288.
9. Zarkin GA, Bray JW, Aldridge A, et al; COMBINE Cost-Effectiveness Research Group. Cost and cost-effectiveness of the COMBINE study in alcohol-dependent patients. Arch Gen Psychiatry. 2008;
10. Zarkin GA, Bray JW, Aldridge A, et al. The effect of alcohol treatment on social costs of alcohol dependence: results from the COMBINE study. Med Care. 2010;48:396-401.
11. Oslin DW, Berrettini W, Kranzler HR, et al. A functional polymorphism on the mu-opioid receptor gene is associated with naltrexone response in alcohol-dependent patients. Neuropsychopharmacology. 2003;28:1546-1552.
12. Anton RF, Oroszi G, O’Malley S, et al. An evaluation of mu-opioid receptor (OPRM1) as a predictor of naltrexone response in the treatment of alcohol dependence: results from the Combined Pharmacotherapies and Behavioral Interventions for Alcohol Dependence (COMBINE) study. Arch Gen Psychiatry. 2008;65:135-144.
13. Weiss RD, O’Malley SS, Hosking JD, et al; COMBINE Study Research Group. Do patients with alcohol dependence respond to placebo? Results from the COMBINE Study. J Stud Alcohol Drugs. 2008;69:878-884.
14. Johnson BA, Rosenthal N, Capece JA, et al; Topiramate for Alcoholism Advisory Board; Topiramate for Alcoholism Study Group. Topiramate for treating alcohol dependence: a randomized controlled trial. JAMA. 2007;298:1641-1651.
Information for professionals about self-help for patients and for the identification and treatment (including medications) of individuals with alcohol use disorders:
• Anton RF. Naltrexone for the management of alcohol dependence. N Engl J Med. 2008;359:715-721.
• National Institute on Alcohol Abuse and Alcoholism. http://www.niaaa.nih.gov/Pages/default.aspx. Accessed March 16, 2011.
For more on the COMBINE study:
• Collaborative Studies Coordination Center. COMBINE Study. http://www.cscc.unc.edu/COMBINE/publications. Accessed March 16, 2011.