• 
• 

Three Part Question

In [patients with alcohol use disorder] does [naltrexone] promote [reductions in alcohol consumption]?

Clinical Scenario

A 49 year old man presents to your family medicine office with signs and symptoms of alcohol use disorder. He is willing to explore pharmaceutical support to reduce his drinking.

Search Strategy

Ovid interface on the world wide web. 1966- April 2020
[naltrexone\tu] AND [alcoholism\ or “alcohol use disorder\”] and [("meta*" or "systematic*").ti,ab,kw,kf.]

Search Outcome

75 papers found, of which 15 were systematic reviews or meta-analyses with at least 250 patients that primarily examined changes in alcohol consumption with naltrexone treatment. 5 further papers included for context in comments section.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Palpacuer, C, et al., 2018 France	1315 patients	Meta-analysis	Total Alcohol Consumption	Standard Mean Difference -0.11, 95% CI -0.40, 0.18	The meta-analysis was attempting to examine a wide range of pharmacological interventions. Though the total population of the meta-analysis was 1315 patients, no individual outcome had more than 750 patients. Consistent results in favour of naltrexone with confidence intervals still overlapping 0 may indicate that the meta-analysis was underpowered to find statistical significance. There was significant heterogeneity in the naltrexone studies, with an I2 of 46.4 - 82.8 depending on outcome.
			Heavy drinking days	Standard mean difference -0.03, 95% CI -0.21, 0.16
			Non drinking days	Standard mean difference -0.28, 95% CI -0.95, 0.40
			Number of drinking days	Standard mean difference -0.16, 95% CI -0.35, 0.04
			Drinks per drinking day	Standard mean difference -0.04, 95% CI -0.31, 0.23
Canidate, S, et al., 2017 United States	903 women with alcohol dependence	Systematic Review	Total alcohol consumption, frequency of alcohol consumption, number of heavy drinking days, number of days abstinent, time to relapse to heavy drinking, time to relapse to any drinking	1 of 7 studies reported statistically significant reduction in drinking. Modest increase in time to relapse and reduction in quantity of drinking among women who took naltrexone vs. placebo.	Studies included were short duration (mode: 8 weeks, max: 12 weeks) and many had very small sample sizes which would only be powered to detect large treatment effects.
Soyka, M, et al., 2016 Germany	974 heavy drinking patients	Indirect Meta-Analysis	Change in baseline drinking frequency	Hedges' g= 0.071, p= 0.62	Indirect meta-analyses have to make statistical choices that can significantly impact outcome. This paper has significant conflict of interest and risk of bias. Indirect meta-analysis was comparing nalmefene to naltrexone. The study was sponsored by Lundbeck, which markets nalmefene. Two of the authors are employees of Lundbeck, and the third has worked as a consultant for Lundbeck for the past 5 years.
			Change in baseline drinking quantity	Hedges' g= 0.108, p= 0.18
Donoghue, K, et al., 2015 United Kingdom	4199 heavy drinking patients	Meta-Analysis	Return to any drinking at 3 months	RR= 0.92, 95% CI= 0.86 - 1.00	Naltrexone had limited studies conducted in Europe, so the finding that there was no change in efficacy may be limited by sample size. Study’s primary outcomes were abstinence related, while abstinence is frequently not the primary goal of treatment in clinical settings.
			Return to heavy drinking at 6 months	RR= 0.85, 95% CI= 0.87-0.93
			Adverse events	No increased total risk of discontinuing treatment with naltrexone compared to placebo, but increased risk due to adverse events
			Geographic variability in treatment effect	No difference in efficacy of naltrexone between Europe and North America
Jonas, DE, et al., 2015 United States	9140 heavy drinking patients	Meta-analysis	Return to any drinking	Risk Difference= -0.05(95% CI -0.10 to -0.002), NNT 20 (95% CI 11 to 500)	Limited and low-quality evidence for naltrexone dosages other than 50mg PO. Significant heterogeneity of trials – some used naltrexone alone, some with other interventions. Populations had heterogenous comorbid medical conditions such as depression. Best evidence related to abstinence, which may not be the primary goal of treatment in clinical settings.
			Return to heavy drinking	Weighted mean risk difference= -0.09 (95% CI -0.13 to -0.04) NNT 12 (95% CI 8 to 26)
			% Drinking Days	Weighted mean risk difference= -5.4 (95% CI -7.5 to -3.2)
			% Heavy Drinking Days	Weighted mean risk difference= -4.1 (95% CI 7.6 to -0.61)
			Drinks per Drinking Day	Weighted mean risk difference= -0.49 (95% CI -0.92 to -0.06)
			Adverse events	Patients treated with naltrexone had higher risk of withdrawal due to adverse events (NNH 48)
Jarosz J, et al., 2013 Poland	2427 heavy drinking adults	Meta-analysis	Alcohol abstinence	OR 1.46 (95% CI 1.07-2.00, p= 0.00182)	It is unclear from the text of the meta-analysis if primary outcomes were defined prospectively. All studies examined naltrexone + psychotherapy, so results may not be applicable to naltrexone without psychotherapy. Individual outcomes had much smaller sample sizes than the total meta-analysis (n= 34 to n= 1,187) and number of studies examining the outcome (1 to 6). Limited evidence beyond 12-16 week period.
			Relapse to Drinking	Odds Ratio 0.48 (95% CI 0.36-0.64, p<.0001)
			Secondary outcomes	Naltrexone also demonstrated superiority to placebo in craving scores, %days abstinent, number of drinks per drinking day, first occurrence of relapse, and GGT levels. Odds Ratio 1.46 (95% CI 1.07-2.00, p= 0.00182) Odds Ratio 0.48 (95% CI 0.36-0.64, p<.0001) Naltrexone also demonstrated superiority to placebo in craving scores, %days abstinent, number of drinks per drinking day, first occurrence of relapse, and GGT levels. No superiority to placebo in GGT level change from baseline, AST, ALT, Accumulated days abstinent, %heavy drinking days, %heavy drinking weeks
Lobmaier, P, et al., 2013 Norway	939 alcohol-dependent patients	Systematic Review	Reduction in drinking	Examined only depot naltrexone injections. Depot injection at 300mg in one study had fewer abstinent days, but no change in relapse to first drinking day compared to placebo.	Only examined two studies because of the limited availability of data on depot injections. Provided limited data about study population other than alcohol dependent status. Both studies included have also been included in several other meta-analyses.
			Relapse to heavy drinking	Depot injections at 380mg showed a 25% reduction in heavy drinking days in one study, and patients at 380mgs reported improved quality of life. That population also showed lower total alcohol consumption among 6 months of follow up. Response to lower dosages were inconsistent and not statistically significant.
Miller, P, et al., 2011 United States	8,493 patients with diagnosed alcohol dependence	Systematic Review	Naltrexone efficacy	Among 15 studies with “less potential for bias” (Jadad scale 4 or 5) which compared naltrexone to placebo, 12 showed statistically significant benefit of naltrexone. Of the 3 negative studies, 2 were specifically targeting populations with comorbid medical conditions. Among 15 studies with high potential for bias, 12 showed statistically significant benefit of naltrexone.	Significant heterogeneity in defining drinking outcomes and patient populations amongst included studies. Some studies compared naltrexone to other active medications rather than placebo. Studies varied broadly in approach to “medical management” and psychosocial support.
			Role of adjunct psychosocial support	Brief psychosocial support may be as effective as more intensive interventions as adjunct to pharmacotherapy
Rosner, S, et al., 2008 Germany	2182 alcohol dependent patients	Meta-analysis	Abstinence	RR= 0.93 (95% CI 0.88 to 0.99) NNT= 17.4 (95% CI 9.1 to 111.0)	Article did not discuss definitions for some of their coding words, such as relapse to heavy drinking. We can assume that the underlying literature has significant heterogeneity, as has been the case for most meta-analyses and systemic reviews. Very limited description of the characteristics of underlying patient population. It may be interesting to comment on the amount of unpublished data from study investigators and drug manufacturers that is included in this meta-analysis, affecting up to 33.8% of the outcome criteria (Table 2)
			Relapse to heavy drinking	RR= 0.80 (95% CI 0.17 to 0.91) NNT= 8.1 (95% CI 5.5 to 16.5)
			Time to first drink	No benefit over placebo Standard Mean Difference = 0.97 (95% CI -0.01 to 1.94), p= 0.052
			Drinking days per week	Standard Mean Difference = -0.14 (95% CI -0.25 to 0.03), p<0.05
			Daily alcohol consumption	No benefit over placebo Standard Mean Difference = -0.89 (95% CI -1.88 to 0.10)
			GGT	Standard Mean Difference = -0.37 (95% CI -0.51 to -0.22), p<0.01
Pettinati HM, et al., 2006 United States	5997 alcohol dependent patients	Systematic Review	Return to any drinking	36% (9 of 25) studies favoured naltrexone over placebo in reducing return to any drinking or increased abstinence.	Studies varied in terms of psychosocial intervention and inclusion of patients with concurrent psychiatric disorders.
			Heavy drinking	70% (19 of 27) studies favoured naltrexone over placebo for reducing heavy or excessive drinking
			Safety	Low frequency of side effects (<15%) Common: Nausea and vomiting Side effects typically resolve spontaneously No instances of hepatotoxicity at 50mg daily
Roozen, HG, et al., 2005 The Netherlands	2,248 patients	Systematic review	Relapse rate	Favors naltrexone. Pooled difference 13% in favour of naltrexone (7% to 18%) NNT = 8	Article was a mixed systematic review and meta-analysis. Though this has several advantages (being able to provide the statistical rigour of a meta-analysis while keeping broad inclusion criteria) it also means that there is a significant difference among the strength of evidence depending on the underlying evidence . Some outcomes where statistical analysis was not possible because a small subset of studies did not provide standard deviations.
			Continuous abstinence	Does not favour naltrexone. Pooled difference 6% (-2% to 15%), p= 0.12
			Time to first relapse	Mixed evidence – 5 of 9 studies favoured naltrexone, 4 of 9 did not. Statistical pooling not possible.
Srisurapanont, M, et al., 2005 Thailand	2,861 patients with diagnosed alcohol dependence or abuse using DSM-III-R or DSM IV	Meta-analysis	Relapse to heavy drinking	RR= 0.64 (95%CI= 0.51 to 0.82) NNT= 7	Included studies without a placebo arm. This study siloed similar results which could result in underpowering (i.e. reporting naltrexone + intensive psychosocial treatment vs naltrexone + simple psychosocial treatment), so for many outcomes this functioned more as a systematic review. Did not include study definitions of relapse to heavy drinking.
			Relapse to any drinking	RR= 0.91 (95%CI= 0.81 to 1.02)
			Safety	Nausea: RR= 2.14 (95%CI= 1.161 to 2.83) NNH= 8 Dizziness: RR= 2.09 (95%CI= 1.28 to 3.39) NNH= 12 Fatigue: RR= 1.35 (95%CI= 1.04 to 1.75) NNH= 17
Carmen B, et al., 2003 Spain	3,205 patients with alcohol dependence using DSM-III-R or DSM-IV	Systematic Review and meta-analysis	Rate of relapse to heavy drinking	OR= 0.62 (95%CI= 0.52 to 0.75)	Secondary outcomes contained relatively few studies with relatively few participants (2-7 studies, 222-1172 participants). This variation means that some outcomes may have been underpowered while others were overpowered.
			Abstinence rate	OR= 1.26 (95%CI= 0.97 to 1.64)
			Safety	Discontinuation due to side effects OR= 2.59 (95%CI= 1.23 to 3.71) p<0.001 Retention rate OR= 0.94 (95%CI= 0.80 to 1.1) p= 0.5
			Secondary outcomes: Primary: Rate of relapse to heavy drinking Abstinence rate Safety Secondary: Time to relapse Time to first drink %drinking days Drinks/drinking day Abstinent days Heavy drinking days Total alcohol consumption (g/week) Liver enzymes (GGT, AST, CDT%) Alcohol cravings	Statistically significant reductions to all secondary outcomes except time to first drink.
Streeton C, et al., 2001 United States	804 heavy drinking adults	Meta-analysis	% Relapse to heavy drinking	Risk difference: -14%	Use of risk difference rather than odds ratio requires a lower bar to reach statistical significance. Authors did not mention conflict of interest, and the associate director of DuPont Pharmaceuticals provided comments to the authors prior to publication. The study seemed to have a single data extractor.
			% Abstinence rate	Risk difference: +10%
			Drinks per drinking day	Average -1.0 drinks
			Adverse events	Naltrexone was no more toxic than placebo, but nausea, somnolence, abdominal pain, anorexia and vomiting were all reported more frequently in treatment arms of studies included. Naltrexone did not have a statistically significant increase in study dropouts than placebo.
Kranzler HR et al., 2001 United States	439-781 patients with alcohol dependence (outcome dependent)	Meta-analysis		All data presented as weighted correlation coefficient r (95% CI)	Small sample size for each outcome relative to other meta-analyses included. Use of r instead of an odds ratio requires lower threshold for statistical significance. Safety was evaluated primarily through retention rates, which is less relevant than adverse events. Authors did not comment on conflicts of interest.
			% Subjects abstinent	r= 0.122 (0.045 to 0.197) p<0.001
			% Drinking Days	r= -0.191 (-0.265 to -0.114) p<0.001
			Drinks/Drinking day	r= -0.067 (0.0161 to 0.028) p=0.081
			% Relapse to heavy drinking	r= -0.161 (-0.243 to -0.077) p<0.001
			Retention %	r= 0.005 (-0.081 to 0.092) p=0.45

Comment(s)

The evidence on naltrexone seems to indicate consistent benefits compared to placebo in a significant number of indicators of drinking behaviour including cravings, length and frequency of alcohol abstinence, and amount of alcohol consumed. The evidence, however, is complicated due to the wide array of different outcomes being measured (Maisel NC et al. 2013), and the relationship of the measures to patient centered outcomes. Naltrexone tends to show different effectiveness profiles for different outcomes (Maisel). Further research would benefit from a clinical consensus on patient-centred goals of treatment, whether that be mortality, quality of life, hospital admissions, reduced total alcohol consumption, reduced heavy drinking etc. The effect was more pronounced in trials lasting at least 12 weeks. The studies presented in this review may somewhat overstate the corpus of evidence; all meta-analyses drew from the same data set of 109 unique RCTs, with the data from Jonas (2005), Roozen (2006), and Carmen (2004) having data sets completely covered in other meta-analyses. Several large studies including Krystal (2001) also represented a significant portion of participants for the meta-analyses, as many studies had fewer than 100 participants in their test arm. A theme among many of the meta-analyses was naltrexone showing greater benefit over placebo in relapse to heavy drinking, but more limited or no benefit in return to any drinking. Many studies involved a lead-in period of abstinence; due to this some of the data about reductions in drinking may have been diluted by patients who remained abstinent (Mason, 2003). There also seems to be a trend towards lower measured effect size in multi-centered trials compared to single centered trials (Feinn R et al., 2005) and over time (Del Re AC, et al., 2013), though it is unclear exactly what is driving the latter effect. There is significant variation among patients’ response to naltrexone, which may be driven by genetic factors (Chamorro AJ et al., 2012), individual etiology of the patient’s alcohol use disorder, comorbid addictions, and comorbid medical conditions. This is a growing area of research. There is variation in dosing for naltrexone, with the most common being 25mg PO daily initial dosing, increased to 50mg PO daily as tolerated. Extended release depot injections are also available as a method of dosing, and can be considered depending on patient preference. Side effects tend to be mild and naltrexone has low risk of serious adverse events that are typically outweighed by the benefits of reduction in drinking, though clinicians should be aware of interactions with NSAIDs, possible hepatic side effects, and causing withdrawal side effects among opioid users (Mason BJ, 2003).

Clinical Bottom Line

There is evidence that naltrexone reduces alcohol cravings and several indicators of drinking behaviour. It has few serious adverse events, though is generally slightly less well tolerated than placebo. Given limited pharmaceutical options, it is a reasonable part of treatment for alcohol use disorders.

Level of Evidence

Level 1 - Recent well-done systematic review was considered or a study of high quality is available.

References

1. Palpacuer C, Duprez R, Huneau A, et al. Pharmacologically controlled drinking in the treatment of alcohol dependence or alcohol use disorders: a systematic review with direct and network meta-analyses on nalmefene, naltrexone, acamprosate. Addiction 2018; 113(2):220-237.
2. Canidate SS, Carnaby GD, Cook CL, Cook RL. A Systematic Review of Naltrexone for Attenuating Alcohol Consumption in Women with Alcohol Use Disorders. Alcoholism: Clinical and Experimental Research 2017; 41(3):466-472.
3. Soyka M, Friede M, Schnitker J. Comparing Nalmefene and Naltrexone in Alcohol Dependence: Are there any Differences? Results from an Indirect Meta-Analysis. Pharmacopsychiatry 2016; 49(2):66-75.
4. Donogue K, Elzebri C, Saunders R, et al. The efficacy of acamprosate and naltrexone in the treatment of alcohol dependence, Europe versus the rest of the world: a meta-analysis. Addiction 2015; 110(6):920-30.
5. Jonas DE, Amick HR, Feltner C, et al. Pharmacotherapy for adults with alcohol use disorders in outpatient settings: a systematic review and meta-analysis. JAMA 2014; 311(18):1889-900.
6. Jarosz J, Miernik K, Wachal M, et al. Naltrexone (50 mg) plus psychotherapy in alcohol-dependent patients: a meta-analysis of randomized controlled trials. American Journal of Drug and Alcohol Abuse 2013; 39(3):144-60.
7. Lobmaier PP, Kunoe N, Gossop M, et al. Naltrexone depot formulations for opioid and alcohol dependence: a systematic review. CNS Neuroscience & Therapeutics 2011; 17(6):629-36.
8. Miller PM, Book SW, Stewart SH. Medical treatment of alcohol dependence: a systematic review. International Journal of Psychiatry in Medicine 2011; 42(3):227-66.
9. Rosner S, Leucht S, Lehert P, et al. Acamprosate supports abstinence, naltrexone prevents excessive drinking: evidence from a meta-analysis with unreported outcomes. Journal of Psychopharmacology 2008; 22(1):11-23.
10. Pettinati HM, O’Brien CP, Rabinowitz AR, et al. The status of naltrexone in the treatment of alcohol dependence: specific effects on heavy drinking Journal of Clinical Psychopharmacology 2006; 26(6):610-25.
11. Roozen HG, de Waart R, van der Windt DA, et al. A systematic review of the effectiveness of naltrexone in the maintenance treatment of opioid and alcohol dependence European Neuropsychopharmacology 2006; 26(6):610-25.
12. Srisurapanont M, Jarusuraisin N. Naltrexone for the treatment of alcoholism: a meta-analysis of randomized controlled trials. International Journal of Neuropsychopharmacology 2005; 8(2):267-80.
13. Carmen, B, Angeles M, Munoz A, et al. Efficacy of naltrexone and acamprosate in the treatment of alcohol dependence: a systematic review. Addiction 2004; 99(7):811-28.
14. Streeton C, Whelan G. Naltrexone, a relapse prevention maintenance treatment of alcohol dependence: a meta-analysis of randomized controlled trials. Alcohol & Alcoholism 2001; 36(6):544-52.
15. Kranzler HR, Van Kirk J. Efficacy of naltrexone and acamprosate for alcoholism treatment: a meta-analysis. Alcoholism: Clinical & Experimental Research 2001; 25(9):1335-41.
16. Maisel NC, Bodgett JC, Wilbourne PL, et al. Meta-analysis of naltrexone and acamprosate for treating alcohol use disorders: when are these medications most helpful?. Addiction 2013; 108(2):275-93.
17. Mason BJ. Acamprosate and naltrexone treatment for alcohol dependence: an evidence-based risk-benefits assessment. European Neuropsychopharmacology 2003; 13(6):469-75.
18. Feinn R, Kranzler HR. Does effect size in naltrexone trials for alcohol dependence differ for single-site vs. multi-center studies? Alcoholism: Clinical & Experimental Research 2005; 29(6):983-8.
19. Del Re AC, Maisel N, Blodgett J, et al. The declining efficacy of naltrexone pharmacotherapy for alcohol use disorders over time: a multivariate meta-analysis Alcoholism: Clinical and Experimental Research 2013; 37(6): 1064-8.
20. Chamorro AJ, Marcos M, Miron-Canelo JA, et al. Association of micro-opioid receptor (OPRM1) gene polymorphism with response to naltrexone in alcohol dependence: a systematic review and meta-analysis. Addiction Biology 2012; 17(3):505-12.