GERALD F. X. FEENEY 1 *, JASON P. CONNOR 2, ROSS MCD. YOUNG 3, JANE TUCKER 1 and ANNIE MCPHERSON 1 INTRODUCTION

Transcription

1 Alcohol & Alcoholism Vol. 41, No. 3, pp , 2006 Advance Access publication 8 February 2006 doi: /alcalc/agl007 COMBINED ACAMPROSATE AND NALTREXONE, WITH COGNITIVE BEHAVIOURAL THERAPY IS SUPERIOR TO EITHER MEDICATION ALONE FOR ALCOHOL ABSTINENCE: A SINGLE CENTRES EXPERIENCE WITH PHARMACOTHERAPY GERALD F. X. FEENEY 1 *, JASON P. CONNOR 2, ROSS MCD. YOUNG 3, JANE TUCKER 1 and ANNIE MCPHERSON 1 1 Alcohol and Drug Assessment Unit, Princess Alexandra Hospital, Brisbane, Queensland 4102, Australia, 2 Department of Psychiatry, The University of Queensland, Brisbane, Queensland 4102, Australia, 3 School of Psychology and Counselling, Faculty of Health, Queensland University of Technology, Queensland 4034, Australia. (Received 14 July 2005; first review notified 19 August 2005; in revised form 3 November 2005; accepted 17 January 2006; advance access publication 8 February 2006) Abstract Aims: To compare treatment outcomes amongst patients offered pharmacotherapy with either naltrexone or acamprosate used singly or in combination, in a 12-week outpatient cognitive behavioural therapy (CBT) programme for alcohol dependence. Methods: We matched 236 patients across gender, age group, prior alcohol detoxification, and dependence severity and conducted a cohort comparison study of three medication groups (CBT+acamprosate, CBT+naltrexone, CBT+combined medication) which included 59 patients per group. Outcome measures included programme attendance, programme abstinence and for those who relapsed, cumulative abstinence duration (CAD) and days to first breach (DFB). Secondary analyses compared the remaining matched 59 subjects who declined medication with the pharmacotherapy groups. Results: Across medication groups, CBT+ combined medication produced the greatest improvement across all outcome measures. Although a trend favoured the CBT+ combined group, differences did not reach statistical significance. Programme attendance: CBT + Acamprosate group (66.1%), CBT + Naltrexone group (79.7%), and in the CBT + Combined group (83.1%). Abstinence rates were 50.8, 66.1, and 67.8%, respectively. For those that did not complete the programme abstinent, the average number of days abstinent (CAD) were 45.07, 49.95, and days, respectively. The average numbers of days to first breach (DFB) was 26.79, 26.7, and days. When the focal group (CBT + combined) was compared with patients who declined medication (CBT-alone), significant differences were observed across all outcome indices. Withdrawal due to adverse medication effects was minimal. Conclusions: The addition of both medications (naltrexone and acamprosate) resulted in measurable benefit and was well tolerated. In this patient population naltrexone with CBT is as effective as combined medication with CBT, but the trend favours combination medication. INTRODUCTION Psychosocial treatments offer definite but limited success in the treatment of alcohol dependence. Two adjunctive medications, acamprosate (calcium acetylhomotaurinate) and naltrexone are of benefit (Bouza et al., 2004). Acamprosate raises the continuous abstinence rate with a calculated number needed to treat (NNT) of 10 [95% confidence interval (95% CI) 7 15] and significantly improves cumulative abstinence both in the treatment and the follow-up period (Bouza et al., 2004; Mann et al., 2004). The evidence for naltrexone is less robust because of smaller study sample sizes and studies of shorter duration (Bouza et al., 2004). However, short-term use (12 weeks or less) significantly reduces relapse rates and has a NNT of 9 (95% CI 6 14) (Bouza et al., 2004). Used separately, acamprosate and naltrexone have both been demonstrated to be safe but their effectiveness is limited because of medication compliance issues. There are insufficient data to establish the superiority of either drug. Medications that reduce alcohol consumption influence several neurotransmitter systems which underlie the reinforcing or discriminative stimulus effects of ethanol and include endogenous opioids, catecholamines, especially dopamine; serotonin, and excitatory amino acids (e.g. glutamate) (Myrick and Anton, 2004). Reports of the clinical benefit of naltrexone, a pure opioid antagonist, in the treatment of alcohol dependence extend back to the early 1990s *Author to whom correspondence should be addressed at: Dr Gerald Feeney, Medical Director, Alcohol and Drug Assessment Unit, Princess Alexandra Hospital, Brisbane, Queensland 4102, Australia, Tel.: ; Fax: ; Gerald_Feeney@health.qld.gov.au (O Malley et al., 1992; Volpicelli et al., 1992). Reduction of the rewarding effects of alcohol by blocking endogenous opioid reward as well as reducing alcohol cue conditioned reinforcement signals is considered the primary mechanism of action. Awareness of the clinical benefits of acamprosate emerged also in the early 1990 s (Lhuintre et al., 1990; Pelc et al., 1992). Acamprosate modulates the activity of the glutamate system (Littleton, 1995). It plays a role in normalizing the dysregulation of the NMDA-mediated glutamergic neurotransmission which occurs with chronic alcohol consumption (Mason, 2003). The precise mechanism by which either of these medications act in alcohol dependence is not yet established. The coadministration of acamprosate and naltrexone significantly increases the rate and extent of acamprosate absorbtion (Mason et al., 2002; Johnson et al., 2003). Whether there is a clinical advantage from this greater systemic availability of acamprosate is unknown. These medications do not modify the acute pharmacological or pharmacokinetic properties of alcohol. Tolerance, dependence or the emergence of a withdrawal syndrome or rebound drinking has not been reported with either medication. Neither naltrexone nor acamprosate has overt psychoactive effects on the central nervous system (Mason, 2003). The efficacy of using these medications in combination with behavioural therapy requires investigation. As both medications act by distinctly different mechanisms there is a rationale for combining both. Naltrexone reduces craving for alcohol. The mechanism through which naltrexone operates may be the extinction of positive reinforcement as it modifies the sense of intoxication from alcohol (Myrick and Anton, 2004). Alcohol withdrawal is associated with reduced GABAergic inhibition and increased glutamatergic excitation. 321 Ó The Author Published by Oxford University Press on behalf of the Medical Council on Alcohol. All rights reserved

2 322 G. F. X. FEENEY et al. As a consequence this reduction of post withdrawal neuronal hyperexcitability by acamprosate may result in reduced physiological and psychological distress and therefore diminish desire for alcohol (Littleton, 1995). The combination may be additive or possibly synergistic. Although clinical benefit with either agent has been known for a decade only recently a landmark study reporting clinical benefit from combining both medications has been conducted. The study, a singlesite, placebo controlled clinical trial was performed in Germany (Kiefer et al., 2003). Naltrexone combined with acamprosate considerably enhanced relapse prevention. Because of this study, combination treatment has become more frequent in clinical practice. The COMBINE study protocol and pilot feasibility study has been published and the findings of the primary study are expected to be published soon. This study examines the efficacy of combination pharmacotherapy interacting with two forms of psychosocial therapy (COMBINE Study Research Group, 2003a,b). As treatment of alcohol dependence is evolving there are no reported Australian data available on combination treatment. In single medication studies we have previously reported improved alcohol abstinence rates over a 12 week cognitive and behavioural therapy (CBT) programme with either naltrexone or acamprosate (Feeney et al., 2001a, 2002). Here we report on our clinical experience by evaluating abstinence outcomes using combination pharmacotherapy compared with the prescription of either naltrexone or acamprosate alone. In addition, recognizing the methodological differences, we compare our findings with the German study. METHODS Subjects Six hundred and three patients were consecutively treated for alcohol dependence. All patients provided written consent to participate in an established cognitive behavioural therapy (CBT) treatment programme. Subjects received pharmacotherapy in the following way: Between July 1998 and February 2000 Naltrexone alone was offered. Acamprosate became listed (tax payer subsidized and available) on the Pharmaceutical Benefits Scheme (PBS) in November In the absence of evidence for the superiority of either medication and as our programme target was alcohol abstinence rather than controlled consumption, Acamprosate alone was offered from February 2000 to February If specific contraindications to Acamprosate occurred (a very unusual occurrence) naltrexone was offered. From February 2004, combined Naltrexone and Acamprosate was exclusively offered, unless contraindications (very unusual) to either agent occurred. At no stage were patients offered a choice between medications. All treatments occurred consecutively and medication was introduced by a single prescriber (GF) across the period of adjunctive pharmacotherapy (July 1998 June 2005). To June 2005, 12.9% were prescribed a combination of naltrexone and acamprosate, 18.2% naltrexone alone and 26.1% acamprosate alone. Almost half (42.8%) accepted CBT but declined adjunctive medication. Data obtained from patients in the last three groups have been employed in previous studies (Feeney et al., 2001a, 2002). A matched design was used. Each subject was individually matched (blind to outcome measures) across the three medication groups and the no medication group on the following parameters: 1. Gender (male/female) 2. Age (0 19, 20 39, 40 59, 60 79) 3. Prior Medically Supervised Detoxification (yes/no) 4. Alcohol Problem Severity (below 50th percentile/above 50th percentile of bmast) Applying this approach from a total of 603 subjects, a sample of 236 subjects, with 59 subjects per treatment group were matched. Where there was more than one possible match, a random selection option was employed using SPSS (V. 11). Patients who could not be paired on the entire individual matching parameters (n = 367, 60.9%) were not included (Fig. 1). There were 38 males and 21 females in each of the four matched treatment groups (n = 152 males, 84 females). Table 1 summarizes demographic and drinking variables between the three medication + CBT groups and includes the CBT-alone group. With the exception of the CBT alone group having lower educational attainment, there were no other significant differences between treatment groups on these parameters. Recognizing a potential bias in the treatment group that declined medication, the primary analyses compared the focal study group (CBT + Naltrexone/Acamprosate combined ) with the other two medication + CBT group (CBT+ Naltrexone and CBT + Acamprosate). An additional secondary comparative analyses included the non-medication group (CBT-alone). Inclusion and exclusion criteria All patients were over the age of 18 years, fulfilled the Diagnostic and Statistical Manual of Mental Disorders (DSM IV) criteria for alcohol dependence and recorded Brief Michigan Alcoholism Screening Test (bmast) scores of >6 (range 0 29 scores, > 6 associated with a high likelihood of alcohol dependence) (Pokorny et al., 1972). All had sustained at least 3 days of alcohol abstinence. Exclusion criteria included current dependence on substances other than alcohol (except nicotine), current disulfiram, prescribed opioid, or psychotropic medication use. Treatment groups Cognitive behavioural therapy treatment programme. All subjects (n = 236) participated in a manualized abstinencebased CBT outpatient programme. The Programme contains defined cognitive and behavioural components conducted by masters and doctoral qualified clinical psychologists. Across the recruitment period (July 1998 June 2005), the same core clinical psychology staff treated all four treatment groups. Where new staff were appointed, all underwent comprehensive training in line with the manualized programme and received weekly supervision from the senior clinical psychologist (R.Y.). Treatment components focus on four core areas: (i) identification and modification of alcohol expectancies, (ii) increasing drinking refusal self-efficacy skills, (iii) developing more effective coping strategies, (iv) teaching problemsolving skills. The programme includes hourly therapy

3 COMBINED EFFICACY OF ACAMPROSATE ABD NALTREXONE 323 Fig. 1. Summary of matching protocol, attendance and programme abstinence outcomes at 12 weeks between CBT-alone, CBT + Acamprosate, CBT + Naltrex- Naltrexone and CBT + Combine groups. Significance difference to focal group (Combine + CBT). * = <0.01, ** = <0.05, # = <0.10 (trend). Table 1. Demographic and drinking history by Treatment Group (n = 59 per group) CBT + naltrexone + acamprosate CBT + acamprosate CBT + naltrexnaltrexone CBT only Measure Mean SD Mean SD Mean SD Mean Mean F P Age Consumption and dependence BMAST Usual consumption (average grams of alcohol per drinking occasion) Maximum consumption (maximum grams of alcohol per drinking occasion) Chronology Age of dependence onset Years of use Other demographic Proportion Proportion Proportion Proportion x 2 P Gender (male proportion) 64.4% 64.4% 64.4% 64.4% Education Primary (Years 1 7) 0% 1.7% 1.7% 9.1% Secondary (Years 8 12) 69.5% 84.5% 71.2% 61.8% Tertiary (College/Adv.) 30.5% 13.8% 27.1% 29.1% sessions weekly for the first 4 weeks. Patients are then seen fortnightly for the subsequent 8 weeks. Medication. All had established at least 3 days abstinence before medication was prescribed. The programme physician (G.F.) discussed adjunctive pharmacotherapy and supplied an information pamplet to all programme participants. For patients who accepted medication, 1 month s supply of medication was prescribed. The physician corresponded with the patient s medical practitioner outlining medication rationale and the need for patients to continue medication over the 12 week CBT programme. Patients were advised to begin medication the day of prescription which was usually before the first CBT session. The interval between the physician assessment and session one of the CBT programme was typically 1 week. Naltrexone (CBT + Naltrexone) subjects (n = 59). In addition to CBT, Naltrexone + CBT subjects were prescribed naltrexone 50 mg orally daily. Acamprosate (CBT + Acamprosate) Subjects (n = 59). In addition to CBT, CBT + Acamprosate subjects weighing >60 kg were prescribed acamprosate calcium 333 mg tablets, two tablets three times daily (1998 mg/day). Those weighing <60 kg received four tablets (1332 mg/day) daily, two tablets in the morning, one at midday and one in the evening.

4 324 G. F. X. FEENEY et al. Combined Acamprosate and Naltrexone (CBT + Combined) subjects (n = 59). In addition to CBT, CBT + Combined subjects received both acamprosate and naltrexone at the above stated dosages. CBT-alone subjects (n = 59). Subjects that declined medication were allocated to the CBT-alone group and were not prescribed adjunctive acamprosate or naltrexone. Secondary analyses were conducted to compare the CBT-alone group with the medication + CBT groups. MEASURES Measures of programme compliance Attendance. The treating psychologist recorded attendance across each of the eight treatment sessions (over 12-weeks) for each patient. Attendance was classified by two outcome indices: (i) number of sessions attended and (ii) completed all session/did not complete all sessions. Abstinence. Alcohol abstinence was estimated by a combination of factors including patient self-report (Alcohol Timeline Follow Back TFLB) (Sobell and Sobell, 1996) and clinic visit alcohol Breathalysation (BAL). Monthly aspartate aminotransferase (AST), Gamma glutamyl transferase (GGT) and carbohydrate deficient transferrin (CDT) estimation (Allen et al., 1999) were used to assess return to heavy drinking. If the patient failed to maintain abstinence, the date of the breach was noted, and days abstinent calculated. If a patient withdrew from treatment, this was recorded as a failure to maintain abstinence. Medication compliance was inquired about on each clinic visit and if medication difficulties arose the patient discussed these with the programme physician. Abstinence was classified by three outcome indices: (i) programme completed abstinent/not abstinent, (ii) cumulative abstinence duration (CAD) (defined as the number of days the non-abstainers remained abstinent over the duration of the 84 day treatment programme) and (iii) days to first breach (DFB) [the number of days until non-abstainers first breached abstinence (1 83 days)]. Statistical analysis The primary analyses compared the focal study group (CBT + Naltrexone/Acamprosate combined ) with the other two medication + CBT groups (CBT + Naltrexone and CBT + Acamprosate). A secondary comparative analyses included the non-medication group (CBT-alone). Chi-square analyses were used to determine if the proportion of programme attendance and abstinence differed among treatment groups. The average days abstinent, CAD and DFB were examined, by group, using ANOVA. Post-hoc tests (i.e. c 2 and Bonferroni) were used to identify differences between the focal group (i.e. CBT + Combined) and other treatment groups. The comparative cumulative proportion of relapse, by treatment group, was assessed using a log rank analysis. All analyses were conducted using SPSS (Version 11). RESULTS Programme attendance by Treatment Group Attended all sessions (programme completion). The programme completion ratio was 39 / 59 in the CBT + Acamprosate group, 47 / 59 in the CBT + Naltrexone group and 49 / 59 in the CBT + Combined group. For the CBT-alone group it was 35 / 59. The difference in the programme completion ratios between the three medication groups reached a statistical trend (c 2 = 5.24, P = 0.073). With the inclusion of the CBT-alone group this reached statistical significance (c 2 = 11.02, P = 0.012). Post-hoc analyses identified significant differences between the focal group (CBT + Com- Combined) and CBT + Acamprosate (c 2 = 4.47, P = 0.035) group. There was also a significant difference between the focal group and the non-medication group (c 2 = 8.09, P = 0.004). Figure 1 summarizes attendance findings. Number of sessions attended. There were no significant differences observed in the average number of sessions attended between the three CBT+ medication groups [F(2, 174) = 2.44, P = 0.090]. When the CBT-only group was included in the analysis, significant differences were observed [F(3, 232) = 5.146, P = 0.002]. Post-hoc analyses revealed the only significant difference was between the focal group (CBT + Combined) and the CBT-alone group (mean difference 1.152, P = 0.006). Figure 1 summarizes attendance findings. Medication tolerability/compliance. Medication complaints were uncommon, and mainly of gastrointestinal origin and usually resolved with temporary dose reduction. Medication withdrawal was reported in <5% of patients and was usually related to intolerance of gastrointestinal discomforts or dysphoria. Among programme non-completers only 6 patients ( 1 Naltrexone alone, 2 Acamprosate alone and 3 Combined medication group) were identified who required medication withdrawal. No biochemical measures of medication compliance were employed. Programme abstinence by Treatment Group Total abstinence. Total programme abstinence was achieved by 30 / 59 of the CBT + Acamprosate group, 39 / 59 for the CBT + Naltrexone group and 40 / 59 for the CBT + Combined group. Alcohol abstinence amongst the non-medication group was 24 / 59. The difference in proportions between the three medication groups was not significant (c 2 = 4.35, P = 0.114). When the CBT-alone group was included in the analysis, there was a significant group difference (c 2 = 12.04, P = 0.007). Post-hoc analyses identified a significant trend for different proportions between the focal group (CBT + Combined) and CBT + Acamprosate (c 2 = 3.51, P = 0.061). The proportions of the combined medication group were also different to the CBT-alone group (c 2 = 8.74, P = 0.003). Figure 1 summarizes programme abstinence findings. Programme participation with relapse to alcohol use Cumulative abstinence duration. There was no significant differences in the average number of days abstinent across the three medication treatment groups for those that did

5 COMBINED EFFICACY OF ACAMPROSATE ABD NALTREXONE Cumulative Abstinence Duration (Days) Fig. 2. Alcohol relapse: cumulative abstinence duration (CAD) by Treatment Group. Significance difference to focal group (Combine + CBT). * = <0.05. not complete the programme abstinent (nonabstainers = CBT + Acamprosate 29 / 59, CBT + Naltrexone, 20 / 59, CBT + Combined 19 / 59 )[F(2, 65) = 0.680, P = 0.510]. When the non-abstinent CBT alone subjects ( 35 / 59 ) were included in the analysis, there was a significant group difference [F(3, 99) = 3.38, P = 0.021]. The CBT + Acamprosate group had a total of days abstinent (95% CI ), CBT + Naltrexone days (95% CI ) and CBT + Combined days (95% CI ). The CBT- alone group had a total CAD of days (95% CI ). Post-hoc analyses revealed a trend for the CBT + Combined group to have higher CAD than the other three groups, however power was restricted due to sample size. Significant differences between the focal group (CBT + Combined) and the CBT-alone group were observed (mean difference 19.67, P = 0.034). Figure 2 (left side of figure) summarizes CAD by group. Days to first breach. If a breach of abstinence occurred, the day of the relapse was noted (range 0 83 days). If a patient withdrew from treatment without advising clinical staff, this was recorded as a relapse. There were no significant differences between groups in the average number of days to the first breach for non-abstaining subjects within the three medication groups [F(2, 65) = 1.95, P = 0.150]. The CBT + Acamprosate group recorded DFB (95% CI ), CBT + Naltrexone 26.7 DFB (95% CI ) and CBT + Combined DFB (95% CI ). The combined group shows a trend for greater DFB. Figure 2 (right side of figure) summarizes DFB by group. When the CBT-alone group was combined in the analysis [16.77 days (95% CI )], a significant difference in DFB was observed [F(3, 99) = 5.57, P < 0.001]. Post-hoc analyses revealed significant differences between the focal group (CBT + Combined) and the CBT- alone group (mean difference 20.55, P < 0.001). A log rank analysis revealed a significant treatment group difference in the onset of first relapse across programme duration (Log Rank = 17.44, df 3, P < 0.001). Post-hoc focal Cumulative Proportion CBT+Combined CBT+Naltrexone CBT+Acamprosate CBT group analyses showed a significant difference between Combined + CBT and Acamprosate + CBT (Log Rank = 4.64, df 1, P = 0.031) and CBT-alone (Log Rank = 13.47, df 1, P = 0.001), but not with CBT + Naltrexone (Log Rank = 0.17, df 1, P = 0.681). A Kaplan Meier survival curve (Fig. 3) shows the comparative cumulative proportion of subjects by treatment group relapsing across the 12-week abstinence based programme. DISCUSSION In alcoholism treatment, effective medication is needed to supplement the modest efficacy of currently available Days to 1st Breach Program Days Fig. 3. The comparative cumulative proportion of relapse during the 12-week abstinence based programme by treatment group.

6 326 G. F. X. FEENEY et al. behavioural treatments. The immediate challenge is to establish abstinence. Although a perception remains that pharmacotherapy does not have a demonstrably large effect on alcoholism treatment, medication continues to be shown to be of value. In this single centre study, the primary goal of pharmacotherapy was to establish and maintain abstinence. Because of the chronic relapsing nature of alcohol dependence, it is useful to assess outcome by the duration of the abstinent interval prior to a lapse and the reduction of drinking intensity if a relapse occurs. The results of this cohort comparison study highlight the progressively better short-term alcohol abstinence rates since the introduction of pharmacotherapy to the CBT programme. The addition of both medications (naltrexone and acamprosate) has resulted in measurable benefit. Although failing to achieve statistical significance, combined medication further improved treatment programme outcomes as measured by attendance and total abstinence rates. Amongst patients who relapsed, combined medication improved the CAD and extended the number of days to first relapse. Combined medication was more effective than both CBT with acamprosate and the group that declined pharmacotherapy (CBT-Alone). Outcome differences between CBT and naltrexone and combined treatment were less well defined. Trend improvement with combined medication was evident in programme completion rates. This was apparent in the abstinence rates, particularly in the first 4 weeks of therapy when psychological treatment was more frequent. Amongst patients who relapsed, those treated with combination medication experienced a longer period of cumulative abstinence. Although combined medication consistently shows better performance than naltrexone alone, we were restricted (in the matched design) by the relatively small sample size. In this patient population naltrexone with CBT is as effective as combined medication with CBT but the trend favours combination medication. Only further larger studies will be able to establish a treatment benefit of combined medication. On the basis of this finding it is possible that the more important clinical effect occurs through a reduction in cue conditioned positive reinforcement signals with a reduction in glutamate related hyperexcitability playing a lesser role. These potential mechanisms require further evaluation. Our findings have both similarities and differences with the German group. Kiefer et al. (2003) performed a double-blind, placebo-controlled study (n = 160). As this study was the first to report the clinical utility of combined medication it has influenced prescribing practice. In addition to methodological differences, psychotherapy in the German study was conducted weekly in the form of group therapy using a CBT model. In spite of these differences we achieved a similar 12 week abstinence rate with combination pharmacotherapy (67.8% vs 65%). Medication withdrawal was low in the Kiefer et al. study at 6.8% and serious adverse reactions did not occur. This was also our experience. The similarity of these findings allowing for the differences in methodology (placebo controlled vs matched historical controls) is noteworthy. Our findings differ across the comparison groups. Compared with the German study, we achieve better abstinence rates across the other pharmacotherapy groups (CBT + naltrexone 66.1% vs 55%, CBT + acamprosate 50.8% vs 42.5%). The disparity between abstinence rates in the CBT alone group in our study and the placebo group in the German study (40.7% vs 25%) may also reflect study design differences. Improvement in abstinence rates across all groups in our study may explain the benefit of combination treatment registering as a trend in the current study compared with the significant difference reported by Kiefer et al. (2003). Our study has several limitations. Randomized controlled trials (RCTs) have important methodological advantages over observational studies including higher internal validity. Selection bias regarding therapy given to an individual patient is overcome in RCTs, however inclusion and exclusion criteria may introduce other types of selection bias and compromise generalizability and representativeness. Our findings are from a single setting, one reason for our comparison with the German study. The potential for base-line differences between patient groups is important as some patients have self-selected to attend our programme others have been referred. Self-selection may have been important factor in the improved outcomes amongst the naltrexone group, as we had access to this medication early and availability was associated with modest media coverage. Treatment assignment bias should have been minimal as all patients were offered a specific medication component, (Naltrexone was offered for 19 months to February 2000, Acamprosate for 48 months to February 2004 and combined medication for 17 months to June 2005). Many declined any adjunctive medication. This reflects clinical reality. The purpose of including this group was to report on our clinical experience of the many patients that make an informed choice not to take adjunctive medication. While there are likely to be fundamental (unmeasured) differences between those who decline a component of treatment, we believe the added ecological validity justifies inclusion as a supplementary analysis. Amongst patients on pharmacotherapy for alcohol dependence, adherence with prescribed medication is an identified problem (Bouza et al., 2004). Medication compliance was not specifically measured although in a previous evaluation, under different prescribing circumstances, it compared favourably with other hospital populations with chronic disorders (Feeney et al., 2001b). Acamprosate and naltrexone appear to act on different behavioural aspects of alcohol dependence which may reflect different neurophysiological actions. This study reflects the way treatments were allocated in actual practice. Combined medication is more effective than either CBT + acamprosate or CBT alone across most major indices and was well tolerated. The safety of combining both medications has been reported (Mason et al., 2002; Johnson et al., 2003) Many promising avenues for pursuit in pharmacotherapy treatment in alcohol dependence treatment remain and include optimal duration of pharmacotherapy, its role in secondary prevention and use in alcohol dependence and co-morbid drug use. REFERENCES Allen, J. P., Sillamaukee, P. and Anton R. (1999) Contribution of carbohydrate deficient transferrin to gamma glutamyl transpepsidase in evaluating progress of patients in treatment for alcoholism. Alcoholism: Clinical and Experimental Research 23,

7 COMBINED EFFICACY OF ACAMPROSATE ABD NALTREXONE 327 Bouza, C., Magro, A., Munoz, A. et al. (2004) Efficacy and safety of naltrexone and acamprosate in the treatment of alcohol dependence: a systematic review. Addiction 99, COMBINE Study Research Group. (2003a) Testing combined pharmacotherapies and behavioural interventions in alcohol dependence. Alcoholism: Clinical Experimental Research 37, COMBINE Study Research Group. (2003b) Testing combined pharmacotherapies and behavioural interventions in alcohol dependence (The COMBINE Study): A pilot feasibility study. Alcoholism: Clinical and Experimental Research 27, Feeney, G. F. X., Young, R. McD., Connor, J. P. et al. (2001a) Outpatient cognitive behavioural therapy programme for alcohol dependence: impact of naltrexone use on outcome. Australian and New Zealand Journal of Psychiatry 35, Feeney, G. F., Connor, J. P., Young, R. McD. et al. (2001b) Adherence with naltrexone prescription advice in hospital outpatient alcohol rehabilitation program. Journal of Clinical Pharmacy and Therapeutics 26, Feeney, G. F. X., Young, R. McD., Connor, J. P. et al. (2002) Cognitive behavioural therapy with relapse prevention medication acamprosate: are short-term treatment outcomes for alcohol dependence improved? Australian and New Zealand Journal of Psychiatry 36, Johnson, B. A., O Malley, S. S., Ciraulo, D. A. et al. (2003) Dose-ranging kinetics and behavioral pharmacology of naltrexone and acamprosate, both alone and combined, in alcohol-dependent subjects. Journal of Clinical Psychopharmacology 23, Kiefer, F., Jahn, H., Tarnaske, T. et al. (2003) Comparing and combining Naltrexone and Acamprosate in relapse prevention of alcoholism. Archives of General Psychiatry 60, Lhuintre, J. P., Moore, N., Tran, G. et al. (1990) Acamprosate appears to decrease alcohol intake in weaned alcoholics. Alcohol and Alcoholism 25, Littleton, J. (1995) Acamprosate in alcohol dependence: how does it work? Addiction 90, Mann, K., Lehert, P. and Morgan, M. Y. (2004) The efficacy of acamprosate in the maintenance of abstinence in alcohol-dependent individuals: Results of a meta-analysis. Alcoholism: Clinical and Experimental Research 28, Mason, B. J. (2003) Acamprosate and naltrexone treatment for alcohol dependence: an evidence-based risk-benefits assessment. European Neuropsychopharmacology 13, Mason, B. J., Goodman, A. M., Dixon, R. M. et al. (2002) A pharmacokinetic and pharmacodynamic drug interaction study of acamprosate and naltrexone. Neuropsychopharmacology 27, Myrick, H. and Anton, R. (2004) Recent advances in pharmacotherapy of alcoholism. Current Psychiatry Reports 6, O Malley, S., Jaffe, S., Chang, G. et al. (1992) Naltrexone and coping skills therapy for alcohol dependence. A controlled study. Archives of General Psychiatry 49, Pelc, I., LeBon, O., Verbanck, P. et al. (1992) Calciumacetylhomotaurine for maintaining abstinence in weaned alcoholic patients: a placebo-controlled double-blind multicenter study. In Novel Pharmacological Interventions for Alcoholism. Naranjo, C. A. and Sellers, E. M. eds, pp Springer Verlag, New York. Pokorny, A. D., Miller, B. A. and Kaplan, H. B. (1972) The brief MAST: a shortened version of the Michigan Alcoholism Screening Test. American Journal of Psychiatry 129, Sobell, L. C. and Sobell, M. B. (1996) Alcohol Timeline Followback (TLFB) Users Manual. Toronto, Canada: Addiction Research Foundation. Volpicelli, J. R., Alterman, A. I., Hayashida, M. et al. (1992) Naltrexone in the treatment of alcohol dependence. Archives of General Psychiatry 49,