Appendix of Naltrexone Implant studies grouped by type.

Study Design (level of evidence) Population Sample

size

(n/group)

Intervention Comparator Period of follow-up  Outcomes measured
RCTs
Kunøe Randomised controlled trial (II) Opioid dependent outpatients who had completed abstinence oriented in-patient treatment and passed an oral naltrexone challenge

(25mg)

56

(29/27)

Go Medical naltrexone implant (2200mg) Usual care- encouraged to contact relevant aftercare such as outpatient counselling, application for entry to the Norwegian maintenance treatment program, re-admission to detoxification, or residential treatment 6 months 1,2,4,5,6,7,8,9,10
Hulse

2009 

 

Randomised controlled trial (II) Opioid dependant outpatients who had completed preclinical screening

 

70

(35/34▪)

 

All participants underwent detoxification followed by

(double placebo controlled): Go Medical naltrexone implant (2300mg) or placebo implant and oral naltrexone tablets (50mg/d) or placebo tablets

Usual care- encouraged to attend weekly individual, group or family therapy 6 months 1,4,5,9
Comparative studies
Carreño

2003

 

Concurrent control

(III-2)

Prospective design

Opioid dependent outpatients 440

(156/284)

ROD using a combination of antagonists followed by long term maintenance with Wedgewood naltrexone implant (1000mg) and CBT Classic detoxification followed by maintenance with oral naltrexone

(dose not reported) and CBT

12 months 3,4,7
Colquhoun

2005

 

Concurrent control  (III-2) Opioid dependent outpatients 83

(41/42)

Go Medical naltrexone implant (1700mg single or double) plus counselling and regular phone support Oral naltrexone (dose not reported) plus counselling and regular phone support 12 months (plus 8 months during treatment) 1,3,7
Grusser

2006

 

Concurrent control  (III-2) Opioid dependent patients 68  (17 in each group) UROD with oral naltrexone, followed by one week of oral naltrexone treatment (50mg/d), followed by naltrexone implant (1000mg brand not stated) 1)                Detoxified and treated with oral Levomethadone (dose not reported)

2)                Actively consuming opioid addicts

3)                Healthy volunteers

12 weeks (compared to data collected 4 days after UROD, and 6 weeks into treatment) 2,8**,10**
Study Design (level of evidence) Population Sample

size

(n/group)

Intervention Comparator Period of follow-up  Outcomes measured
Hulse

2005a^

 

Cohort- pre/post test (IV). Record linkage Opioid dependent outpatients

 

361

 

Go Medical naltrexone implant (2200mg corrected by Ngo 2007) simultaneously with ROD 6 months pre treatment 6 months

 

1,4,5
Ngo

2008a^

 

Cohort- pre/post test (IV). Record linkage Opioid dependant outpatients 359 Go Medical naltrexone implant (2200mg) simultaneously with ROD Pre treatment period truncated to match post treatment follow-up 1.78 years post-treatment 8
Ngo

2008b^

 

Concurrent cohort

(III-2)

Retrospective record linkage

Opioid dependent outpatients 836

(314/

522)

Go Medical naltrexone implant (2200mg) simultaneously with ROD MMT (methadone syrup prescribed by medical practitioners and dispensed by pharmacists), no information on dosage 3.5 years (plus 6 months pre

treatment)

 

1,4,5
Ngo 2007^

 

Case series- pre/post test  (IV). Record linkage Opioid dependent outpatients 359 Go Medical naltrexone implant (2200mg) simultaneously with ROD Pre treatment period truncated to match post treatment follow-up 1.78 years post-treatment 8
Reece

2007

 

Concurrent and historical control study, within group

comparisons pre/post test on outcomes of interest

(IV). Record linkage

Opioid dependent outpatients 376

(102/113/

161)

Go Medical and Wedgewood naltrexone implants (1100mg and 1000mg respectively).  25 patients (24.5%) had multiple implants Revia naltrexone tablets (50mg/d) a) concurrent control

b) historical control (12 months prior to NIG and above)

12 months 2,5,7
Tait

2008a^

 

Cohort- pre/post test (IV). Record linkage Opioid dependent outpatients who had received oral naltrexone prior to implant 130 Go Medical naltrexone implant (2200mg) simultaneously with ROD 6 months pre and post first selfreported heroin use 6 months pre and post oral naltrexone

6 months pre and post implant

6 months 1,8
Tait

2008b^

 

Concurrent cohort  (III-2) Record

linkage

 

Opioid dependent outpatients 894

(341/

553)

Go Medical naltrexone implant (2200mg) simultaneously ROD MMT 3.5 to 5.5 years 4
Study Design (level of evidence) Population Sample

size

(n/group)

Intervention Comparator Period of follow-up  Outcomes measured
Waal 2006

 

Case series- pre/post test (IV) Opioid dependent outpatients 13 Go Medical naltrexone implant (1800mg single, double or single + double) 1 week pre treatment 11.3 months 1*,4, 5*,7*,8, 9,10
Studies with no control group
Foster

2003

 

Consecutive

cohorts - post

test  (IV)

Opioid dependent

patients

 

101

(55/46)

Wedgewood implant

(1000mg single, double or single + double) 1st cohort: ROD under general anesthesia followed by implant

2nd cohort: ROD (non-iv sedation) followed by implant

N/A

No comparison made between cohorts

12 weeks 1,4,10
Olsen

2004

 

Case series- post test  (IV) Opioid dependent outpatients enrolled in professional treatment or counselling programs 10 Wedgewood naltrexone implant (1000mg single, 3 or 4 implants) N/A 80 days 4,9*
ROD, Rapid Opioid Detoxification; UROD, Ultrarapid Opioid Detoxification; CBT, Cognitive Behavioural Psychotherapy

Notes:

1) Opioid use;

2) Treatment adherence;

3) Retention in treatment: time to drop out;

4) Adverse effects;

5) Use of drugs other than opioids;

6) Criminal activity and incarceration;

7) Quality of life;

8) Mental health;

9) Duration of achieved therapeutic naltrexone blood levels;

10) Heroin craving

* Not included in analysis of this outcome as there was no control group

** Not included in analysis of this outcome because measurements were taken four days after detoxification to assess its effectiveness for opioid detoxification ^ Same base cohort

▪1 participant indicated that they were not heroin dependant so were excluded from analysis.

APPENDIX C: CHARACTERISTICS OF EXCLUDED STUDIES

 

 

Study Reason for exclusion
Farid 2008 Not a clinical trial (review)
Gibson 2007 Sample size less than 10
Hall 2008a Not a clinical trial (review)
Hall 2008b Not a clinical trial (comment)
He 2009 Outcome measures outside of protocol (memory ability). No adverse effect data reported
Hulse 2008 Outcome measures outside of protocol (biodegradability). No adverse effect data reported
Hulse 2005b Outcome measures outside of protocol (pharmacodynamic results on tissue biopsies)
Hulse 2004a Sample size less than 10. No adverse effect data reported
Hulse 2004b Outcome measures outside of protocol (maternal and neonatal outcomes). No adverse effect data reported
Hulse 2003 Sample size less than 10. No adverse effect data reported
Hulse 2002a Sample size less than 10. No adverse effect data reported
Hulse 2002b Sample size less than 10. No adverse effect data reported
Lintzeris 2008 Sample size less than 10 (total sample size was 12, but <10 participants reported as opioid dependent)
Montoya 2008 Not a clinical trial (review)
Sullivan 2007 Intervention other than naltrexone implants (oral naltrexone)

30

APPENDIX D: QUALITY ASSESSMENT OF INCLUDED STUDIES

 

RCTs

Study Random

assignment to treatment groups

Blinding at treatment allocation Method of blinding adequately described Eligibility criteria described Groups comparable at baseline Groups treated identically apart from intervention Reasons for withdrawal described and ITT used Outcomes measured in a reliable way
Kunøe

2009

 

Y

Computer generated

randomisation

list

N

Open

allocation of

participants

 

Y

Open label envelopes sealed and

numbered

by staff independent of the study and provided to participants as per

randomised

list

Y

Implantation and participation free

to all participants. Reasons for ineligibility: not completing inpatient treatment, awaiting transfer to other clinics and starting maintenance treatment.  High degree of motivation

Y

Group differences analysed using two-way ANOVA with

Bonferroni correction. The only significant pre-treatment difference between groups was benzodiazepine use, which was

controlled for

 

N

Usual care uncontrolled, statistical analysis used to control for possible confounding factors (except the influence of treatment centre on measures of mental health

and quality of life)

 

Y

93% followup at 6 months. ITT analysis performed. The worst recorded score in the group was used for missing responses that lacked an equivalent item at inclusion

U Y Y

Self reports: Addiction Severity Index, timeline follow-back interview technique, DSM-IV diagnoses using the Mini

International

Neuropsychiatric

Interview (MINI), Euro-

ASI, Beck Depression

Inventory, Hopkins

Symptom Checklist,

Temporal Satisfaction With Life Scale. Heroin use verified against hair samples

N
Hulse 2009 Y

Computer generated randomisation codes

Y

Packs were labelled with randomisation codes. Personnel generating codes and handling medications

Y Study research officers undertaking assessment, recruitment and follow up did not have access Y Opioid dependant, residing in Perth, 18+, willingness to be randomised. Self reported screening questionnaire and Y

Baseline characteristics not significantly different. Data were collected on age, gender, body weight, and opiate

Y

All participants underwent detoxification prior to treatment allocation. Both groups received an

Y

6 (2 ONG, 4 NIG) withdrew without heroin use. 4 stayed in contact, 2 unavailable to follow up.

Y Y Y

Self reported data on opioid and other drug use and adverse events, opioid or other drug overdose, any other treatments.

Blood naltrexone levels measured using liquid chromatography-mass

Y
Study Random

assignment to treatment groups

Blinding at treatment allocation Method of blinding adequately described Eligibility criteria described Groups comparable at baseline Groups treated identically apart from intervention Reasons for withdrawal described and ITT used Outcomes measured in a reliable way
did not have contact with participants to the codes medical examination.

Exclusion criteria

≥3 opioid overdoses in past month, oral naltrexone treatment > 3 times in past 4 months, previous sustained release naltrexone treatment, enrolled in other opioid dependence research, pregnancy, active infection, contraindications to naltrexone,

inability to complete study protocol

 

treatment Index

Social Functioning and General Health

Questionnaire

implant (active naltrexone or placebo) and tablets (active naltrexone or placebo). Both groups were encouraged to attend weekly individual, group or family therapy Overall 5

ONG and 4 NIG

unavailable for follow up. 1 participant contracted active

MRSA and

was switched

from NIG to ONG on medical advice. ITT analysis performed- missing blood naltrexone levels were classified as

0

 

spectrometry.

Unsupervised urine drug screening

Y, yes; N, no; U, unclear; MMT, methadone maintenance treatment; ITT, intention to treat analysis; ONG, oral naltrexone group; NIG , naltrexone implant group; ^ Same base cohort

 

Comparative studies

 

Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
Carreño

2003

 

Y

First 156 patients treated under the NIMROD program (ROD using a combination of antagonists, followed by naltrexone implant plus individual motivational orientation and CBT. NIG: 98% males, mean age 28.8 years, 75% had a partner, 83.3% stable work, 79.1% middle to technical education level. 80.58% used heroin, 6.08% used methadone, and 13.33% used heroin and methadone, 14.6% injected

Y

Outpatients with

criteria of opioid dependence (ICD-10), compliance with the centres admission criteria, no concomitant disease, contraindications or conditions making follow-up difficult, willing to consent and started program between

September 1998 and

October 2000

 

N

Pre treatment, the NIG scored higher than the oral group on occupation/ support (p<0.05), no adjustments reported

U

U if included in analysis Y Reported: 101/156 (65%) completed the program. Of these, only 56% returned for 6 month follow up and 21% for 12 month follow up

U U Y

EuropASI and retention index (assessed objectively).

EuroASI assesses psychiatric status, family/social relationships, legal status, drug use, alcohol use, occupation/ support and medical status

N
Colquho

un 2005

 

Y

Opioid dependent outpatients. NIG: 25 (61%) males, 16 (39%) females.

Mean age 26.2 years, mean years using opiates 7.2 (sd 5.0), mean years of education 10.6 years. A large number of participants in both groups were moderately to highly depressed

Y

Clients had to be motivated to be opiate free, have suitable social support and no psychiatric diagnoses

or medical issues that

would make

detoxification dangerous

N

No significant difference (p<0.05) between groups in age, years using opioids, years of education, daily opioid dose, psychiatric symptoms, and depression. Patients self selected treatment group

Y N

Number of counselling sessions was not significantly different between the

groups

 

N N

Self-esteem and quality of relationships using a 10 point Likert scale. Both administered via telephone survey of participants and their support person

U
Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
Grusser

2006

 

N

Small sample. NIG: 3 females and 14 males, mean age 34.76 years (sd 9.63). The mean duration of heroin use across all four groups was 7.02 years

(sd 5.82)

N N

In part. Age, duration of heroin use and gender were considered but patients in groups one and two chose their treatment

N

Reported but not included in analysis at 6 and

12 weeks

U U U

Unclear how relapse rates were assessed. Drug dependence: DSM-IV-TR

Craving and current mood: Questionnaire on

Differentiated

Assessment of Addiction Depression:

General

Depression Scale

 

U
Hulse 2005a^ Y

Opioid dependent persons who had not previously been treated with a naltrexone implant. 218 (60%) male, 143 (40%) female. Mean duration of heroin use 5.7 years (sd 5.3). 174 (48%) prior oral naltrexone maintenance, with 17% having entered into oral naltrexone maintenance >3 times. Males were significantly older than the females [mean 28.5, sd 7.2 vs mean 26.6 sd 7.9; t 2.4, p

<0.017]

Y

Opioid users treated for the first time with a naltrexone implant at a not for profit

community based clinic in Perth between Jan 2001 and Dec 2002. 361 (94%) of 384 possible participants considered eligible,

i.e. could be followed up via the WALD. Exclusion: no consent given (5), implants removed in the first week (3), and incarceration prior to implant (15)

N

In part. Age, age of first heroin use, duration of heroin use, gender, previous use of oral naltrexone considered. Mentioned confounder of age and sex in relation to mortality

Y

1 participant died and 1 had an early removal, both were included in the analysis. 326 (90%) were in the WALD, 257 (71%) were in the EDIS and 335 (93%) were in both

U Y

Prospectively collected data: assessor blind to treatment

allocation at outcome

assessment

 

Y

Prospectively collected hospital admission and mortality data

from the WALD and the EDIS. Overdoses were identified using

ICD10 codes

N/A

due to record linkage design

Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
Ngo

2008a^

Y

As per Hulse 2005a

Y

As per Hulse 2005a

Y

Potential confounders such as age, gender, preexisting history of mental illness and length of heroin use were

considered

 

Y

1 participant died

U Y

Prospectively collected data: assessor blind to treatment allocation at outcome

assessment

 

Y

Prospectively collected hospital admission and mortality data from the WALD  cases identified using ICD-9-CM diagnosis codes

Y

2 cases with missing data were removed from analysis

Ngo

2008b^

Y

Base cohort as per Hulse

2005a

 

NIG: 314; 129 (41%) females

[mean age 27.1 years sd 8.1],

185 (59%) male [mean age

29.0 years sd 7.5]

 

MMT: 522; 198 (37.9%) females [mean age 30.4 years sd 9.1 years],

324 (62.1%) males [mean age

32.1 years sd 8.9 years]

Y

As per Hulse 2005a

 

MMT cohort

ineligible if taking methadone for pain management, treatment crossover, or incarcerated  

 

N

Age and sex and their modifying

interactions were examined but other factors, e.g. level of motivation, situational influences, socioeconomic background, and pre existing

illness were not

 

N

82.6% MMT, 92.2% NIG lost to follow up, not included in analysis

U The MMT

cohort was only included as a reference group, so between group comparisons must be viewed with great caution

Y

Hospital morbidity and mortality collected prospectively and independent of research

team

 

Y

Prospectively collected hospital morbidity data

from the

WALD.

Physician coded mental health hospital admission using ICD 9 or 10. Inconsistencies in coding between different ICD codes are inevitable

N/A

due to record linkage design

Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
Ngo 2007^ Y

As per Hulse 2005a

Y

As per Hulse 2005a

Y

Investigated influence of gender, age, prior history of hospital admission, and duration of heroin use. Pre treatment truncated to match post

treatment

 

U

Not reported. Likely to be as

per Hulse 2005a

 

U Y

Hospital morbidity and mortality collected prospectively and independent of research

team

 

 

Y

Physician coded mental health hospital admission using ICD-9 and 10, but use of diagnostic codes from different versions means coding inconsistencies

are inevitable

 

N/A

due to record linkage design

Reece

2007

 

Y

Patients presenting to a low cost private naltrexone clinic in Brisbane.

NIG:  69 (68%) male, 33(32%) female. mean age 26.1 years, mean duration of opiate use 5.5 years. 9% on social security alone, 64% in socioeconomic class III-V and

27% in socioeconomic class III. 25% had ‘weak’ social support, 51% ‘average’ and

25% ‘strong’

N

No information on types of participants eligible for programs offered by the clinic

N

Matched for age, sex, duration of opioid use and dose used. Participants selected treatment. Financial status and social support better in the NIG, differences not

controlled for

 

Y

Overall followup was 82%. Urine drug screen followup was 76%. ITT analysis performed

U U N

Primary outcome was ‘opiate free success’, defined as a composite measure including self report, having a urine drug screen negative for opiates or having a carer report satisfactory progress

U
Tait

2008a^

Y

Base cohort as per Hulse 2005a, but restricted to treatment resistant group. Mean history of heroin dependency 7.5 years (sd 4.9),

Y

As per Hulse 2005a had to have received oral naltrexone 12 months prior to their implant, leaving

U Y

As per Hulse

2005a  1 participant died

Y 24 (19%) received an agonist therapy post implant, Y

Mortality data collected prospectively and independent

Y

ICD-9 or 10

N/A

due to record linkage design

Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
which is higher than average. 59% male, 41% female. 45% had only ever received oral and implantable naltrexone, 34% reported use of agonist treatment prior to implant treatment

 

130/139 (93.5%) mainly MMT, dates are unavailable of research

team

 

Tait

2008b^

Y

Base cohort as per Hulse

2005a

NIG: 40% female [mean age

27.0 years sd 7.9], 60% male

[mean age 28.8 years; sd 7.4]

 

Y

As per Hulse 2005a

Y

Only for age and gender

Y

As per Hulse

2005a

U

Additional treatment episodes possible but not reported

Y

Mortality data collected prospectively and independent of research team

Y

WALD which compiles extensive health information for individuals

N/A

due to record linkage design

Waal 2006 Y

Actively seeking this type of treatment. Small sample, 11 male, 2 female [mean age 26.9 sd 4.9]. Mean history of heroin dependency 4.8 years (sd 3.3). Most had extensive experience

with several other psychoactive substances. Participants had received a range of previous treatments

Y

Participants were actively interested in naltrexone implants

U

 

Y

2 participants interviewed with retrospective data

U N/A

due to pre/ post test deign

Y

Liquid chromatography/ mass spectrometry operated in electrospray ionisation, Hopkins

symptom

checklist, Beck depression rating scale, side effects checklist, Likert scales

 

N

Compliance with monthly evaluation procedures varied. Only 7/13 participants accounted for on mental health and craving outcomes at week 12

Y, yes; N, no; U, unclear; MMT, methadone maintenance treatment; NIG naltrexone implant group; ^ Same base cohort; ROD, rapid opioid detoxification; WALD, West

Australian Health Services Research Linked Database

Studies with no control group

 

Study Was the sample representative of patients in the population as a whole? Were recruitment/ eligibility criteria reported? Were losses to follow-up reported and included in analysis? Were losses to follow up> 20%? Were outcomes measured in a reliable way? Were missing data (group or time-point data) accounted for?
Olsen 2004 Y

Small sample. 5 male, 5 female [median age 30.5 (23-29)]. Median heroin use of 7.5 years (4-15). All HCV positive. Under no obligation to abstain from drug use

 

Y

Participants had to be taking part in a professional treatment or counselling program

Y

 

N

3 participants were lost to follow up. Adverse effect results reported for all participants

U

 

N/A

No missing data

Foster 2003 Y

Opioid dependent outpatients, seeking treatment in private clinic.

Cohort one: 76% male, 24% female [mean age 29.8 years sd 5.57]. 51% unemployed, 33% current and 67% previous injectors, mean opiate habit

6.7 years

Cohort two: 83% male, 17% female [mean age 27.4 years]. Comprehensive demographic data not collected

 

N Y N N

Self report (semi structured interview be telephone) with corroboration by family members/carers where possible. Craving on a 10 point visual analogue scale. Blood concentrations measured in several patients not in these cohorts

N/A
Y, yes; N, No; U, unclear; HCV, hepatitis C virus

 

APPENDIX E: RESULTS OF STUDIES ON EFFECTIVENESS

 

1)     Opioid use during and after treatment

 

RCTs

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Kunøe

2009

 

Norway

56

(29/27)

Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant 2200mg Usual care, i.e. encouraged to contact relevant aftercare: outpatient counselling, Norwegian maintenance treatment program, re admission to detoxification, or residential treatment 6 months Between group comparison

a)                6 month follow up: heroin use significantly lower in NIG (mean 17.9 days) vs controls (mean 63.6 days) difference 45.6 p<0.05, F 7.0 [95% CI 14.1, 77.3] ASI 30 day variable, NIG mean 3.5 days of heroin use vs 11.4 days in controls, difference 8, p<0.05, F 5.8 [95% CI 1.8, 14], and ASI frequency scale, difference 0.73, p<0.05 [95% CI 0.11, 1.34]

b)                6 month follow up for all opioid use, NIG scored significantly lower

(mean 37 days) than controls (mean 97.1 days) difference 60.2 p<0.01,

F 8.1 [95% CI 20.9, 99.5] ASI 30 day variable, NIG mean 6.3 days of opioid use vs 17.4 days in controls, difference 9, p<0.01, F 5.4 [95% CI 1.6, 16.4], and ASI frequency scale, difference p<0.001 [95% CI 0.45, 1.7].

c)                Abstinence from all opioids for the whole period was not significant, 11/29 in the implant group vs 5/27 in controls.

d)                At the 6 month follow-up assessment opioid dependence as per DSM-IV diagnoses using MINI was significantly lower in the implant group (9/29) vs controls (18/27), OR= 0.225, p=0.015 [95% CI 0.07, 0.69]

Note: The results of hair analysis matched self-reported opioid use in patients available for testing (86%).

 

Hulse

2009

 

Australia

70 (35/34)

 

Opioid dependant outpatients who had completed preclinical screening, exclusion criteria detailed Go Medical naltrexone

implant (2.3g)

 

Oral naltrexone

tablets

(50mg/d)

All underwent detoxification and were implanted with active or placebo implant.  All were encouraged to attend weekly individual, group or family therapy 6 months

 

Between group comparison

At 6 month follow-up, heroin use was significantly lower in the implant group

[hazard ratio 4.49; 95%CI 1.85,10.90]

Return to regular heroin use occurred significantly earlier in the oral group (median 115 days SE 12.0) compared to the implant group (median 158 days SE 9.4) p=0.001.

12% and 6% of oral and NIG, respectively, returned to heroin 1 to 3 days per week.

0% and 14% of oral and NIG, respectively, returned to heroin 1 to 3 days per month. Self reported complete abstinence in 27% and 63% of the oral and implant groups respectively.

 

NHMRC Literature Review: Naltrexone Implants for Opioid Dependence                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                39

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Colquhou

n 2005

 

Australia

83

(41/42)

Opioid dependent outpatients Go Medical naltrexone

implant (1700mg

single or double) plus counselling and regular phone support

Oral naltrexone plus counselling and regular phone support 12 months (plus 8 months during

treatment)

 

Between group comparison

Abstinent participants were those who had used opioids only a few times since detoxification.

 

In patients who had remained abstinent from opioids, relapse in the NIG was higher than control at 6 months post treatment, 33/41 (80%) vs 23/42 (55%) respectively. This trend continued at 12 months post treatment, where rates were 25/41 (61%), vs 17/42 (40%) respectively. No ratio or CI reported.

 

In patients who had relapsed from opioids (including those non contactable) relapse in the NIG was lower than controls at 6 months post treatment, 8/41(20%) vs 19/42(45%) respectively. This trend continued at 12 months post treatment, where rates were 16/41 (39%), vs 25/42 (60%) respectively. No ratio or CI reported

 

Hulse

2005a^

 

Australia

 

361

 

Opioid dependent

outpatients

 

Go Medical naltrexone implant (2200mg) simultaneously

with ROD

 

6 months pre treatment 6 months

 

Within group comparison

For emergency department presentations and hospital admissions and when data from emergency department presentations and hospital admissions were combined, there were more opioid overdoses 6 months pre treatment vs 6 months post treatment but the significance was not reported. Refer to Ngo 2008b

 

Ngo

2008b^

 

Australia

 

836

(314/ 522)

Opioid dependent outpatients Go Medical naltrexone implant (2200mg) simultaneously with ROD MMT

(methadone syrup prescribed by medical practitioners and dispensed by pharmacists)

3.5 years (plus

6 months pre treatment)

 

Between group comparison

There was no significant change in risk for other opioid hospitalisations from 6 months pre treatment to 6 months post treatment for either cohort. In the MMT cohort there was no significant change in risk of other opioid hospitalisations pre treatment to 3.5 years post treatment whereas for the implant cohort it was significantly reduced, 181 events vs 332 events respectively,

OR 0.64 [95% CI 0.46, 0.89], α = 0.05

There was a significant decline in hospital admission rates from opioid overdose from 6 months pre treatment to 3.5 years post treatment in the NIG,

RR 0.29 [95% CI 0.15, 0.55], α = 0.05 whereas the change was not significant in the MMT cohort.

There was a significant increase in hospital admission rates for other opioid conditions from 6 months pre treatment to 3.5 years post treatment in the MMT cohort, RR 1.35 [95% CI 1.14-1.61], α = 0.05 vs a significant decline in the NIG RR 0.55 [95% CI 0.46-0.65], α = 0.05. In the NIG female patients 30 years and older incurred fewer other opioid hospital admissions than their younger counterparts, RR 0.32 [95% CI 0.16-0.65], α = 0.05

 

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Reece

2007

 

Australia

376

(102/ 113/

161)

Opioid dependent outpatients Go Medical and

Wedgewood

naltrexone implants (1.1g and 1g respectively) Many had multiple implants

Revia naltrexone

tablets

(50mg/d)

a)                concurrent control

b)                historical control (12 months prior to NIG)

12 months

 

Within group comparison

Within the NIG social support, work prior to treatment and the presence of cannabis in the urine were significantly related to opiate free success.

Tait

2008a^

 

Australia

130 Opioid dependent

outpatients who had received oral naltrexone prior to implant

Go Medical naltrexone implant (2200mg) 6 months pre and post: first heroin use, oral naltrexone treatment, implant 6 months Within group comparison

There was a significant decline in opioid overdoses from 6 months before oral naltrexone to 6 months post implant treatment (Wilcoxon Z= 2.3, p=0.02).

 

Opioid overdoses increased from 0 in both the 6 months before and after first heroin use, to 7 in the both the 6 months before and after oral naltrexone to 2 in the 6 months before implant and 0 in the 6 months post implant (Friedman 2 15.2 (5), p =0.01).

MMT, methadone maintenance treatment; NIG, naltrexone implant group; ROD, rapid opioid detoxification; ^ same base cohort

 

2) Treatment adherence

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent

outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg)

Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare: outpatient

counselling, Norwegian maintenance treatment program, re admission to detoxification, or residential treatment

6 months Between group comparison

At 6 months follow up there was no significant difference in outpatient treatment attendance between the implant group and controls, p>0.05.

 

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Grusser

2006

 

Germany

68

(17/17 )

Opioid dependent patients UROD with

oral naltrexone, followed by one week of oral naltrexone treatment (50mg/d), followed by naltrexone implant (1000mg)

2)                Detoxified and treated with oral

Levomethadon

e

3)                Actively consuming

opioid addicts  4) Healthy volunteers

12 weeks compared with data collected 4 days after detoxificat ion, and 6 weeks into

treatment

 

Within group comparison

a)  6 weeks after treatment:

There was one relapse in the NIG.

Only 4 participants remained in group 2; 10 had relapsed and 3 had switched to long-term Levomethadone without any intention of becoming drug free. No significance reported.

 

b)  12 weeks after treatment

2 participants in the NIG had relapsed. None of the remaining 4 subjects in group 2 could be located. No significance reported.

 

Reece

2007

 

Australia

 

376 (102/

113/ 161)

Opioid dependent outpatients Go Medical and

Wedgewood

naltrexone implants (1.1g and 1g respectively), many had multiple implants

Revia naltrexone

tablets

(50mg/d)

a)                concurrent control

b)                historical control (12 months prior to NIG and

above)

 

12 months

 

Between group comparison

The proportion of patients in each group whose urine tested positive for naltrexone were 70%, 52% and 39% respectively.

NIG, naltrexone implant group; UROD, ultrarapid opioid detoxification; ^ same base cohort

 

3)  Retention in treatment: time to drop out

 

Comparative studies

Study/ Country N (n/group ) Population Intervention Comparator Follow-up Main Findings
Carreño

2003

 

Spain

440

(156/ 284)

Opioid dependent outpatients UROD

combination of antagonists

followed by Wedgewood

naltrexone implant

(1000mg) and

CBT

 

Classic detoxification followed by maintenance with oral naltrexone and

CBT

12 months Between group comparison

At 6 and 12 months post treatment the treatment retention index was significantly greater in the implant group (80% at 6 months, 65% at 12 months) than the oral group (42% at 6 months, 17% at 12 months), p<0.05. No CI reported.

Colquhou

n 2005

 

Australia

 

83

(41/ 42)

Opioid dependent outpatients Go Medical naltrexone

implant (1700mg

single or double) plus counselling and regular phone support

Oral naltrexone plus counselling and regular phone support 12 months (plus 8 months during

treatment)

 

Abstinent participants were those who had used opioids only a few times since detoxification.

 

Within group comparisons

In the NIG the time compliant to naltrexone was greater in the abstinent group than those who had relapsed at 6 and 12 months, p<0.05. NIG participants abstinent at 6 and 12 months estimated the implant was effective for 6 months. Those NIG participants who relapsed estimated the implant was effective for 4 months, p<0.05.

 

Between group comparisons

In patients who had remained abstinent from opioids, time spent in treatment was higher in the NIG than the ONG at 6 months post treatment, mean 176.6 days (sd 68.1) vs mean 120 days (sd 104.8) respectively. No p value or CI reported. This trend remained at 12 months post treatment [mean 187.3 days (sd 69.1) vs 123.21 days (sd 105.5) respectively]. No p value or CI reported.

 

In patients who had relapsed to opioid use, time spent in treatment was higher in the

NIG than the ONG at 6 months post treatment, [mean 112.5 days (sd 50) vs mean 19.7 days (sd 31.7) respectively]. No p value or CI reported. This trend remained at 12 months post treatment [mean 120 days (sd 45.6) vs 30.1 days (sd 54.25) respectively]. No p value or CI reported.

 

NIG, naltrexone implant group; ONG, oral naltrexone group; UROD, ultrarapid opioid detoxification

5)  Use of drugs other than opioids during and after treatment

RCTs

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Between group comparison

At 6 month follow up there was no significant difference in self reported use of alcohol or non opioid drugs between the NIG and controls, p>0.05.

 

Hulse

2009

 

Australia

70 (35/34)

 

Opioid dependant outpatients who had completed preclinical screening,  exclusion criteria detailed

 

Go Medical naltrexone implant

(2.3g)

 

Oral naltrexone

tablets

(50mg/d)

All underwent detoxification and were implanted with an active or placebo implant. All participants were encouraged to attend weekly individual, group or family therapy.

 

6 months Between group comparison

Overall use of non opioid drugs was similar between groups [hazard ratio 0.58 95% CI 0.32, 1.05]

 

Most frequently reported category of non opioid drug was cannabis, followed by stimulants.

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Hulse

2005a^

 

Australia

361

 

Opioid dependent

outpatients

 

Go Medical naltrexone implant (2200mg) simultaneously with ROD 6 months pre treatment 6 months

 

Within group comparison

There were more sedative overdoses in the 6 months post treatment than the 6 months pre treatment as measured by emergency department presentations and hospital admissions, significance not reported. Refer to Ngo 2008b

 

The number of other drug overdoses was the same in the 6 months pre treatment as the 6 months post treatment, as measured by emergency department presentations.

There were more other drug overdoses in the 6 months post treatment than the 6 months pre treatment as measured by hospital admissions, significance not reported.

 

When hospital admissions and emergency department presentations were combined, there were less sedative overdoses in the 6 months pre treatment (8, 1.9%) than the 6 months post treatment (16, 4.4%) p <0.01. Nine overdoses were in the first 10 days after treatment and if excluded the trend is neutral or reversed. Other drug overdoses increased from the 6 months pre treatment (2, 0.6%) to post treatment (5, 1.4%), significance not reported.

 

Ngo

2008b^

 

Australia

836  (314/

522)

Opioid dependent outpatients Go Medical naltrexone implant (2200mg) simultaneously with ROD MMT

(methadone syrup prescribed by medical practitioners and dispensed by pharmacists)

3.5 years (plus 6 months pre

treatment)

 

Between group comparison

Comparing non opioid overdose admissions from 6 months pre treatment to

6 months post treatment, no significant change in either cohort for patients aged

25 years, significantly increased risk in patients aged 35 years. MMT: OR 5.03 [95 % CI 1.18, 21.54], α = 0.05

NIG: OR 16.31 [95% CI 3.07, 86.53], α = 0.05

There was no significant change in risk of non opioid oversdose admissions from 6 months pre treatment to 3.5 years post treatment for either cohort.

 

Within group comparison

MMT: no significant change in risk of other non opioid admissions 6 months pre treatment to 6 months or 3.5 years post treatment

NIG: significantly increased risk at both follow-up periods,

RR 1.33 [95% CI 1.01, 1.73], α = 0.05. 12  vs 31 events at 6 months OR 2.54 [95% CI 1.19, 5.43], α = 0.05 and 105 vs 138 events at 3.5 years, OR 1.52 [95% CI 1.04, 2.23], α = 0.05

In the NIG reduced risk for each one year age increment, at 3.5 years the OR was 0.96 [95% CI 0.94, 0.98].

 

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Reece

2007

 

Australia

376

(102/ 113/

161)

Opioid dependent outpatients Go Medical and

Wedgewood

naltrexone implants (1.1g and 1g respectively), many had multiple implants

Revia naltrexone

tablets

(50mg/d)

a)   concurrent control

b)   historical control (12

months prior to NIG and

above)

 

12 months

 

Between group comparison

Urinary analysis was performed to test use of amphetamines, morphine,

Tetrahydrocannabinol (THC), cocaine and methadone. The rate of obtaining a urine drug screen for each of the groups was 98%, 70% and 76% respectively.

Amphetamine use was significantly greater in the NIG than either the concurrent or historical ONG or compared to both controls combined,  Chi Square 7.13, df 1, OR 2.78 [95% CI 1.22, 6.42], p=0.0075 Chi Square 11.14, df 1, OR 3.28 [95% CI 1.52, 7.18], p<0.001

Chi Square 13.83, df 1, OR 3.05 [95% CI 1.60, 5.83], p=0.0002 respectively

 

MMT, methadone maintenance treatment; NIG, naltrexone implant group; ONG, oral naltrexone group; ROD, rapid opioid detoxification; ^ same base cohort

 

6)  Criminal activity and incarceration

 

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Between group comparison

At 6 months follow up there was no significant difference in self reported criminal activity between the implant group and controls, p>0.05.

 

7)            Quality of life

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Between group comparison

a)                Life satisfaction: at 6 month follow up NIG scored higher than controls on the

Temporal Satisfaction With Life Scale, mean difference 5.4, p<0.05, F 5.3 [95% CI

0.68, 10.1]

b)                Recommend treatment to a friend: at 6 month follow up NIG scored higher than controls, mean 85 vs 56.5 respectively, on a visual analogue scale of treatment satisfaction, mean difference 28.5, p<0.05, F 9.1 [95% CI 10.8, 46.2].

c)                Satisfaction with treatment allocation: at 6 month follow up NIG scored higher than controls, mean 78 vs 36 respectively, mean difference 42, p<0.01, F 25 [95% CI 25.9, 58.5]

 

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Carreño

2003

 

Spain

440

(156/ 284)

Opioid dependent outpatients ROD using a combination of antagonists followed by long term maintenance Wedgewood

naltrexone implant

(1000mg) and

CBT

 

Classic detoxification followed by maintenance with oral

naltrexone and

CBT

12 months Between group comparison

No significant differences were found between the two groups at either follow-up, except at 6 months post treatment the NIG scored better than the ONG on family/social relationships, p= 0.05, no CI reported

 

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Colquhoun

2005

 

Australia

83

(41/ 42)

Opioid dependent outpatients Go Medical naltrexone implant

(1700mg  single or double) plus counselling and regular phone support

Oral naltrexone plus counselling and regular phone support 12 months (plus 8 months during

treatment)

 

Abstinent participants were those who had used opioids only a few times since detoxification.

 

Within group comparisons: in participants who were successful in abstaining from opioid use there was a significant increase in self esteem and quality relationships between pre treatment and 6 and 12 months post treatment, p<0.01, in both groups. No CI reported.

Between group comparisons: in participants successful in abstaining from opioid use, the differences between the NIG and the ONG on measures of self esteem and relationship quality at pre treatment were not significant, p=0.86 and 0.81 respectively. In this same group at 6 months post treatment, the NIG had lower self esteem ratings than the ONG (mean difference 1.2), p=0.018. No CI reported. The ONG also tended to score better on relationships at 6 months post treatment compared to the NIG, but the difference was not significant, p=0.055.

Among abstinent participants, there was no significant difference in self esteem or relationship quality at 12 months post treatment, p>0.05

 

Reece

2007

 

Australia

376

(102/ 113/

161)

Opioid dependent outpatients Go Medical and Wedgewood

naltrexone implants (1g and 1.1g respectively), many had multiple

implants

 

Revia naltrexone

tablets, (50mg/d)

c)   concurrent control

d)   historical control (12 months prior to NIG and

above)

 

1 month

for this aspect of the study (plus 1 month before

ROD)

 

Within group comparison

Within the NIG and concurrent ONG work status improved from before to after treatment,

NIG: 18% pre treatment vs 50% after treatment, weighted OR 4.67 [95% CI 2.35, 9.33], p<0.001

Concurrent ONG: 25% pre treatment vs 48% post treatment

weighted OR 2.78 [95% CI 1.52, 5.09], p<0.001. The improvement in the historic ONG was not significant, p=0.178

NIG, naltrexone implant group; ONG, oral naltrexone group; ROD, rapid opioid detoxification; CBT, cognitive behavioural psychotherapy; ^ same base cohort

8)  Mental health

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Between group comparisons

At 6 month follow up there was no significant difference in self reported depression (Beck Depression Inventory and subscale of the 25-item Hopkins Symptom Checklist) between the implant group and controls, p>0.05.

 

 

 

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Ngo

2008a^

 

Australia

359 Opioid dependant outpatients Go Medical naltrexone implant (2200mg) simultaneously with ROD Pre treatment period truncated to match post treatment follow-up 1.78 years posttreatment Within group comparison

Counts of patients and hospital events was categorised into non-substance related, mood related, substance related or all-cause.

 

1. Comparison of pre- versus post-treatment:

a) risk of patient hospitalisation for a mental condition 

The number of cases for patients after naltrexone treatment was consistently lower than pre-treatment. However there was not a statistically significant difference. b) rate of mental health related hospital admissions

There was a reduction in mental health related hospital events post naltrexone treatment. There was a significant reduction in admissions for non-substance mental disorders IRR .630 [95% CI .478, .841 p=.0017]. This was mostly observed in male patients (p=.0002) compared to females (p=.5363). There was a significant reduction in admissions for substance-related mental disorders IRR .673 [95% CI .558, .811 p<.0001] and All-cause mental disorders IRR .641 [95%

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
CI .536, .766 p<.0001]

 

2. Evaluating contributing factors to risk of patient hospitalised due to mental illness following treatment:

Higher risk:

-  of hospitalisation for NSRMD with history of hospitalisation for NSRMD OR

4.78 [95% CI 2.53, 9.01 p<.0001]. Females at higher risk OR 2.39 [95% CI

1.25, 4.58 p.0085].

-  >5 years of heroin dependence OR 3.82 [95% CI 1.08, 13.5 p.0378].

-  of hospitalisation for substance related mental disorder with history of hospitalisation for substance related mental disorder OR 1.82 [95% CI 1.10, 3.04 p.0208]. Females at higher risk OR 1.95 [95% CI 1.21, 3.15 p.0063].

 

3. Length of stay

a)        median per episode varied with different time points in the younger cohort, effect not seen in older patients.

b)        total per person year increased by 1.6 days post treatment, which is similar to the ‘during use’ period.

 

Ngo

2007^

 

Australia

359 Opioid dependent outpatients Go Medical naltrexone implant (2200mg) Pre treatment period truncated to match post treatment follow-up 6 months Within group comparison

The number of hospital admissions significantly decreased pre to post treatment for NSRMD [RR 0.630 95% CI 0.472, 0.841, p = 0.0017], substance related mental disorder [RR 0.673 95% CI 0.558, 0.811, p<0.0001], and all cause mental disorder [RR 0.641 95% CI 0.536, 0.766, p<0.0001]. The number of hospital admissions decreased pre to post treatment for mood disorders, but after adjusting for confounding effects the trend was not significant (p= 0.4232), declining mainly among males [RR 0.463 95% CI 0.220- 0.974] and participants treated at a later age (26 years or older), [RR 0.356 95% CI 0.190-0.668].

There was no significant difference in risk of hospitalisation pre vs post treatment for NSRMD [OR 1.309 95% CI 0.877, 1.954], mood disorders [OR 1.283 95%

CI 0.689, 2.390], substance related mental disorders [OR 1.214 95%

CI 0.886, 1.664], or all cause mental disorders [OR 1.272 95% CI 0.931, 1.738]. Higher risk of non substance related hospital event post treatment with prior hospitalisation for NSRMD OR 4.78 [95% CI 2.53, 9.01]. Females were at higher risk than males, OR 2.39 [95% CI 1.25- 4.58].

 

Participants with >5 years of heroin use pre treatment were at higher risk from hospitalisation due to a mood disorder post treatment compared to those who had used for a shorter period, OR 3.82 [95% CI 1.08, 13.5].

Participants with a history of substance-use hospital diagnoses were at higher risk

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
from hospitalisation due to substance use disorder post treatment OR 1.82 [95% CI 1.10, 3.04]. Females were at higher risk than males, OR 1.95 [95% CI 1.21, 3.15]. Older age at treatment was a protective factor for not developing a drug-related problem post treatment OR 0.949 (0.949- 0.958 for every one year age increment).

Females were at greater risk of all cause mental disorders than males, OR 1.87 [95% CI 1.16, 3.01]. Older age at treatment was a protective factor, OR 0.951 (0.942 0.960 for every one year age increment).

Length of hospital stay due to mental disorders was not significantly different from pre to post treatment, p=04120.

 

Tait

2008a^

 

Australia

130 Opioid dependent

outpatients who had received oral naltrexone prior to implant

Go Medical naltrexone implant (2200mg) 6 months for each: pre and post first heroin use,  Pre and post oral naltrexone, pre and post implant 11.3 months Within group comparison

General trend for an increase in mental health admissions with an opioid related diagnoses in the 6 months post first heroin use, pre and post oral naltrexone periods, with a decline in the 6 months pre and post implant (Friedman65.2 (5), p <0.001). There was a significant decline in mental health admissions with an opioid related diagnoses from 6 months before oral naltrexone to 6 months post implant treatment (Wilcoxon Z = 2.3, p<0.0001).

The trend in mental health admissions with a drug related diagnosis other than opioids was not significant over the 6 study periods (Friedman9.9 (5) p =0.082). In the 6 months pre implant there were two non opioid mental health admissions, 1 with stimulants and 1 with stimulants plus cannabis. In the 6 months post implant there 5 non opioid mental health admissions, 2 with unspecified multidrug use, and 1 each for sedatives, stimulants and cannabis use.

The trend in mental health admissions excluding any drug related diagnoses was not significant over the 6 study periods (Friedman3.6 (5) p =0.616).

 

Waal

2006

 

Australia

13 Opioid dependent outpatients Go Medical naltrexone implant (1800mg

single, double or single + double)

1 week pre treatment 11.3 months Within group comparison

Anxiety/stress levels remained relatively stable pre and post treatment, [ mean 1.9 (SD 0.69) pre treatment vs 1.3 (SD 0.32) to 1.7 (SD 0.58) post treatment] P values not reported.

 

Depression levels generally decreased from 1 week pre treatment to 8 weeks post treatment, [ mean 16.8 (SD 13.8) pre treatment vs 7.6 (SD 7.2) to 12.0 (SD 12.1) in the 8 weeks post treatment]. Depression scores increased at the end of the 12 weeks post treatment[ mean 14.7 (SD 10.4)] P values not reported.

 

NSEMD, non substance related mental disorder; ^same base cohort

9)  Duration of achieved therapeutic naltrexone blood levels

RCTS

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Within group comparison

Plasma levels of naltrexone stayed above 1ng/ml for 6 months and above 2ng/ml for about 5 months. Levels of 6-β-naltrexone had a similar distribution.

 

Note: only 14 of 29 (48%) participants presented for testing of plasma levels.

 

Hulse

2009

 

Australia

70 (35/34)

 

Opioid dependant outpatients who had completed preclinical screening, exclusion criteria detailed Go Medical naltrexone implant

(2.3g)

 

Oral naltrexone

tablets

(50mg/d)

All underwent detoxification and were implanted with an active or placebo implant.  All participants were encouraged to attend weekly individual, group or family therapy.

 

6 months Between group comparisons

Significantly more of the ONG had blood naltrexone levels <2ng/mL at months 1 (ONG 72% NIG 6%) and 2 (ONG 81% NIG 48%) and <1ng/mL at months 1 through 4 (month 1: ONG 59% NIG 0%; month 2: ONG 74% NIG 18%; month 3: ONG 80% NIG 53%; month 4: ONG 97% NIG 53%)

 

Blood naltrexone concentration estimates Men (mean Body Mass Index 25.5, age 33):

≥ 2ng/ml 56 days [95% CI 39, 73]

≥ 1ng/ml 110 days [95% CI 83, 119]

 

Women (mean BMI 25.1, age 28):

≥ 2ng/ml 43 days [95% CI 16, 79]

≥ 1ng/ml 124 days [95% CI 88, 175]

 

 

Comparative study

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Waal

2006

 

Australia

13 Opioid dependent outpatients Go Medical naltrexone

implant (1800mg

single, double or single +

double)

 

1week pre treatment 11.3 months Within group comparison

For participants without measurable naltrexone before implantation, plasma levels were maintained above 1-2ng/ml between 1 and 3 months after a single implant and 3 and 5 months after a double implant. The 6-β-naltrexone concentrations were approximately 1-2.5 times higher compared to naltrexone during the whole implant period.

 

10)  Heroin craving

 

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent

outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg)

Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for

entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential treatment.

 

6 months Between group comparisons

At 6 month follow up the implant group scored lower than controls on a visual analogue scale of craving related thoughts,

mean difference 27, p<0.01, F 7.2 [95% CI 9.4, 44.2]

 

 

Comparative study

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Waal

2006

 

Australia

13 Opioid dependent outpatients Go Medical naltrexone

implant (1800mg

single, double or single +

double)

 

1 week pre treatment 11.3 months Within group comparison

On a scale of 1 (not at all) to 7 (very strong). craving dropped from pre treatment to

1-8 weeks post treatment [mean 2.6 (SD 1.0) pre treatment vs 2.4 (SD 1.6) to

2.7 (SD 1.4) post treatment]. Craving increased at week 12, when the possibility of relapse came closer [mean 3.1 (SD 2.1)], p values not reported.

 

APPENDIX F: RESULTS OF STUDIES ON SAFETY

4)  Adverse Effects

 

RCTS

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, Norwegian maintenance treatment program, re admission to detoxification, or residential treatment. 6 months a)                Site related: 2 participants reported wound opening with leakage of fluid, 3 had allergic reactions all of which were treated. 3 participants had implants removed (1 due to necrosis, 2 on request due to discomfort- site pain or diarrhoea).

b)                Possibly naltrexone related: 1 participant in the NIG overdosed using a combination of opioids, amphetamines and benzodiazepines, 4non fatal over doses were reported in control group.

c)                Mortality: 1 in NIG prior to implant due to overdose, 1 in control group due to overdose 3 months into study.

Hulse

2009

 

Australia

70 (35/34)

 

Opioid dependant outpatients who had completed preclinical screening, exclusion criteria detailed Go Medical naltrexone implant

(2.3g)

 

Oral naltrexone

tablets

(50mg/d)

All underwent detoxification and were implanted with an active or placebo implant.  All participants were encouraged to attend weekly individual, group or family therapy. 6 months a)                Site related: 1 participant had a severe adverse event related to the surgical procedure– wound hematoma proximal to the implant site.

b)                Possibly naltrexone related: 10 study related unexpected adverse events (4 oral, 6 implant) – typically diarrhoea or headaches.

29 study related expected adverse events (10 oral, 19 implant) – typically opiate withdrawal symptoms (diarrhoea, nausea, vomiting), exudation redness, pain at the surgical incision site or proximal to the implant in the first 14 days after surgery.  No opiate overdoses requiring hospital treatment or admission were reported during the 6 month follow up period.

NIG, naltrexone implant group

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Carreño

2003

 

Spain

440

(156/ 284)

Opioid dependent outpatients ROD using a combination of antagonists followed by long term

maintenance Wedgewood

naltrexone implant

(1000mg) and

CBT

 

Classic detoxification followed by maintenance with oral

naltrexone and

CBT

12 months a) Site related: 7 (4.5%) had local allergic tissue reactions. 3 (1.9%) had local wound infection, but no implants had to be removed.

 

Hulse

2005a^

 

Australia

361 Opioid dependent

outpatients

 

Go Medical naltrexone implant (2200mg) simultaneously

with ROD

 

6 months pre treatment 6 months

 

a)                Site related: 3 implant removals in the first week at the patients request, 2 for psychological reasons and 1 due to infection at the wound site. Another removal at 169 days post implant due to an allergic reaction.

b)                Mortality: 1 death due to head trauma, coroners report not available

Ngo

2008b^

 

Australia

836

(314/ 522)

Opioid dependent outpatients Go Medical naltrexone implant (2200mg) simultaneously

with ROD

 

MMT

(methadone syrup prescribed by medical practitioners and dispensed by

pharmacists)

 

3.5 years (plus 6 months pre treatment)

 

a)                Site related: as per Hulse 2005a

b)                Mortality: 1 drug related death in the NIG vs 6 in the MMT group. In the MMT group 2 deaths were opioid related, 1 was a combined drug overdose, and 2 were non-opioid drug overdoses. The mortality in the NIG was from an opioid overdose in a participant previously treated with MMT.

Tait

2008b^

 

Australia

894

(341/ 553)

Opioid dependent outpatients Go Medical naltrexone implant (2200mg) MMT 3.5 to 5.5 years a)  Site related: as per Hulse 2005a

b)  Mortality: 6/341 (1.8%) versus 15/553 (2.7%).

The age standardized mortality rate ratio for naltrexone compared to methadone was 0.645 [95% CI 0.123, 1.17].

The mortality rates for the initial 14 day period’, ‘stable treatment (0-6 months) and overall’ were 0.0, 4.21 and 3.76 vs 94.47, 0.0 and 5.83/1,000 p-y.

In the NIG 2 deaths were drug related and 1 self inflicted. In the MMT group 5 cases were drug related and 2 were self-inflicted.

 

Waal

2006

 

Australia

13 Opioid dependent outpatients Go Medical naltrexone implant (1800mg

single, double or single + double)

1 week pre treatment 11.3 months a)                Site related: 2 local tissue reactions, 1 was mild and responded to antihistamines, the other (of which the patient had had 4 earlier implants) developed necrosis and required removal. At 12 weeks post treatment 9 could still feel their implant on palpitation and 6 felt some anxiety or discomfort by the lump.

b)                Possibly naltrexone related: Side effects (on a checklist including cephalgia, nausea, diarrhoea, muscle and articular pain or discomfort, anxiety and irritability) were recorded for all 13 participants. In the first week pre treatment, diarrhoea, muscle pain and irritability were the most frequency reported vs muscle pain, irritability and anxiety in the first week post treatment. The mean value of all side effect scores was in the range of 'not at all' to 'small' throughout the post implant observation period.

MMT, methadone maintenance treatment; NIG, naltrexone implant group; ROD, rapid opioid detoxification; CBT, cognitive behavioural psychotherapy; ^ same base cohort

Studies with no control group

Study Design (level of evidence) Population Sample

size

(n/group)

Intervention Comparator Period of follow-up  Outcomes measured
RCTs
Kunøe

2009

 

Randomised controlled trial (II) Opioid dependent outpatients who had completed abstinence oriented in-patient treatment and passed an oral naltrexone challenge

(25mg)

56

(29/27)

Go Medical naltrexone implant (2200mg) Usual care- encouraged to contact relevant aftercare such as outpatient counselling, application for entry to the Norwegian maintenance treatment program, re-admission to detoxification, or residential treatment 6 months 1,2,4,5,6,7,8,9,10
Hulse

2009 

 

Randomised controlled trial (II) Opioid dependant outpatients who had completed preclinical screening

 

70

(35/34▪)

 

All participants underwent detoxification followed by

(double placebo controlled): Go Medical naltrexone implant (2300mg) or placebo implant and oral naltrexone tablets (50mg/d) or placebo tablets

Usual care- encouraged to attend weekly individual, group or family therapy 6 months 1,4,5,9
Comparative studies
Carreño

2003

 

Concurrent control

(III-2)

Prospective design

Opioid dependent outpatients 440

(156/284)

ROD using a combination of antagonists followed by long term maintenance with Wedgewood naltrexone implant (1000mg) and CBT Classic detoxification followed by maintenance with oral naltrexone

(dose not reported) and CBT

12 months 3,4,7
Colquhoun

2005

 

Concurrent control  (III-2) Opioid dependent outpatients 83

(41/42)

Go Medical naltrexone implant (1700mg single or double) plus counselling and regular phone support Oral naltrexone (dose not reported) plus counselling and regular phone support 12 months (plus 8 months during treatment) 1,3,7
Grusser

2006

 

Concurrent control  (III-2) Opioid dependent patients 68  (17 in each group) UROD with oral naltrexone, followed by one week of oral naltrexone treatment (50mg/d), followed by naltrexone implant (1000mg brand not stated) 1)                Detoxified and treated with oral Levomethadone (dose not reported)

2)                Actively consuming opioid addicts

3)                Healthy volunteers

12 weeks (compared to data collected 4 days after UROD, and 6 weeks into treatment) 2,8**,10**
Study Design (level of evidence) Population Sample

size

(n/group)

Intervention Comparator Period of follow-up  Outcomes measured
Hulse

2005a^

 

Cohort- pre/post test (IV). Record linkage Opioid dependent outpatients

 

361

 

Go Medical naltrexone implant (2200mg corrected by Ngo 2007) simultaneously with ROD 6 months pre treatment 6 months

 

1,4,5
Ngo

2008a^

 

Cohort- pre/post test (IV). Record linkage Opioid dependant outpatients 359 Go Medical naltrexone implant (2200mg) simultaneously with ROD Pre treatment period truncated to match post treatment follow-up 1.78 years post-treatment 8
Ngo

2008b^

 

Concurrent cohort

(III-2)

Retrospective record linkage

Opioid dependent outpatients 836

(314/

522)

Go Medical naltrexone implant (2200mg) simultaneously with ROD MMT (methadone syrup prescribed by medical practitioners and dispensed by pharmacists), no information on dosage 3.5 years (plus 6 months pre

treatment)

 

1,4,5
Ngo 2007^

 

Case series- pre/post test  (IV). Record linkage Opioid dependent outpatients 359 Go Medical naltrexone implant (2200mg) simultaneously with ROD Pre treatment period truncated to match post treatment follow-up 1.78 years post-treatment 8
Reece

2007

 

Concurrent and historical control study, within group

comparisons pre/post test on outcomes of interest

(IV). Record linkage

Opioid dependent outpatients 376

(102/113/

161)

Go Medical and Wedgewood naltrexone implants (1100mg and 1000mg respectively).  25 patients (24.5%) had multiple implants Revia naltrexone tablets (50mg/d) a) concurrent control

b) historical control (12 months prior to NIG and above)

12 months 2,5,7
Tait

2008a^

 

Cohort- pre/post test (IV). Record linkage Opioid dependent outpatients who had received oral naltrexone prior to implant 130 Go Medical naltrexone implant (2200mg) simultaneously with ROD 6 months pre and post first selfreported heroin use 6 months pre and post oral naltrexone

6 months pre and post implant

6 months 1,8
Tait

2008b^

 

Concurrent cohort  (III-2) Record

linkage

 

Opioid dependent outpatients 894

(341/

553)

Go Medical naltrexone implant (2200mg) simultaneously ROD MMT 3.5 to 5.5 years 4
Study Design (level of evidence) Population Sample

size

(n/group)

Intervention Comparator Period of follow-up  Outcomes measured
Waal 2006

 

Case series- pre/post test (IV) Opioid dependent outpatients 13 Go Medical naltrexone implant (1800mg single, double or single + double) 1 week pre treatment 11.3 months 1*,4, 5*,7*,8, 9,10
Studies with no control group
Foster

2003

 

Consecutive

cohorts - post

test  (IV)

Opioid dependent

patients

 

101

(55/46)

Wedgewood implant

(1000mg single, double or single + double) 1st cohort: ROD under general anaesthesia followed by implant

2nd cohort: ROD (non-iv sedation) followed by implant

N/A

No comparison made between cohorts

12 weeks 1,4,10
Olsen

2004

 

Case series- post test  (IV) Opioid dependent outpatients enrolled in professional treatment or counselling programmes 10 Wedgewood naltrexone implant (1000mg single, 3 or 4 implants) N/A 80 days 4,9*
ROD, Rapid Opioid Detoxification; UROD, Ultrarapid Opioid Detoxification; CBT, Cognitive Behavioural Psychotherapy

Notes:

1) Opioid use;

2) Treatment adherence;

3) Retention in treatment: time to drop out;

4) Adverse effects;

5) Use of drugs other than opioids;

6) Criminal activity and incarceration;

7) Quality of life;

8) Mental health;

9) Duration of achieved therapeutic naltrexone blood levels;

10) Heroin craving

* Not included in analysis of this outcome as there was no control group

** Not included in analysis of this outcome because measurements were taken four days after detoxification to assess its effectiveness for opioid detoxification ^ Same base cohort

▪1 participant indicated that they were not heroin dependant so were excluded from analysis.

APPENDIX C: CHARACTERISTICS OF EXCLUDED STUDIES

 

 

Study Reason for exclusion
Farid 2008 Not a clinical trial (review)
Gibson 2007 Sample size less than 10
Hall 2008a Not a clinical trial (review)
Hall 2008b Not a clinical trial (comment)
He 2009 Outcome measures outside of protocol (memory ability). No adverse effect data reported
Hulse 2008 Outcome measures outside of protocol (biodegradability). No adverse effect data reported
Hulse 2005b Outcome measures outside of protocol (pharmacodynamic results on tissue biopsies)
Hulse 2004a Sample size less than 10. No adverse effect data reported
Hulse 2004b Outcome measures outside of protocol (maternal and neonatal outcomes). No adverse effect data reported
Hulse 2003 Sample size less than 10. No adverse effect data reported
Hulse 2002a Sample size less than 10. No adverse effect data reported
Hulse 2002b Sample size less than 10. No adverse effect data reported
Lintzeris 2008 Sample size less than 10 (total sample size was 12, but <10 participants reported as opioid dependent)
Montoya 2008 Not a clinical trial (review)
Sullivan 2007 Intervention other than naltrexone implants (oral naltrexone)

NHMRC Literature Review: Naltrexone Implants for Opioid Dependence                                              30

APPENDIX D: QUALITY ASSESSMENT OF INCLUDED STUDIES

 

RCTs

Study Random

assignment to treatment groups

Blinding at treatment allocation Method of blinding adequately described Eligibility criteria described Groups comparable at baseline Groups treated identically apart from intervention Reasons for withdrawal described and ITT used Outcomes measured in a reliable way
Kunøe

2009

 

Y

Computer generated

randomisation

list

N

Open

allocation of

participants

 

Y

Open label envelopes sealed and

numbered

by staff independent of the study and provided to participants as per

randomised

list

Y

Implantation and participation free

to all participants. Reasons for ineligibility: not completing inpatient treatment, awaiting transfer to other clinics and starting maintenance treatment.  High degree of motivation

Y

Group differences analysed using two-way ANOVA with

Bonferroni correction. The only significant pre-treatment difference between groups was benzodiazepine use, which was

controlled for

 

N

Usual care uncontrolled, statistical analysis used to control for possible confounding factors (except the influence of treatment centre on measures of mental health

and quality of life)

 

Y

93% followup at 6 months. ITT analysis performed. The worst recorded score in the group was used for missing responses that lacked an equivalent item at inclusion

U Y Y

Self reports: Addiction Severity Index, timeline follow-back interview technique, DSM-IV diagnoses using the Mini

International

Neuropsychiatric

Interview (MINI), Euro-

ASI, Beck Depression

Inventory, Hopkins

Symptom Checklist,

Temporal Satisfaction With Life Scale. Heroin use verified against hair samples

N
Hulse 2009 Y

Computer generated randomisation codes

Y

Packs were labelled with randomisation codes. Personnel generating codes and handling medications

Y Study research officers undertaking assessment, recruitment and follow up did not have access Y Opioid dependant, residing in Perth, 18+, willingness to be randomised. Self reported screening questionnaire and Y

Baseline characteristics not significantly different. Data were collected on age, gender, body weight, and opiate

Y

All participants underwent detoxification prior to treatment allocation. Both groups received an

Y

6 (2 ONG, 4 NIG) withdrew without heroin use. 4 stayed in contact, 2 unavailable to follow up.

Y Y Y

Self reported data on opioid and other drug use and adverse events, opioid or other drug overdose, any other treatments.

Blood naltrexone levels measured using liquid chromatography-mass

Y

NHMRC Literature Review: Naltrexone Implants for Opioid Dependence                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  31

Study Random

assignment to treatment groups

Blinding at treatment allocation Method of blinding adequately described Eligibility criteria described Groups comparable at baseline Groups treated identically apart from intervention Reasons for withdrawal described and ITT used Outcomes measured in a reliable way
did not have contact with participants to the codes medical examination.

Exclusion criteria

≥3 opioid overdoses in past month, oral naltrexone treatment > 3 times in past 4 months, previous sustained release naltrexone treatment, enrolled in other opioid dependence research, pregnancy, active infection, contraindications to naltrexone,

inability to complete study protocol

 

treatment Index

Social Functioning and General Health

Questionnaire

implant (active naltrexone or placebo) and tablets (active naltrexone or placebo). Both groups were encouraged to attend weekly individual, group or family therapy Overall 5

ONG and 4 NIG

unavailable for follow up. 1 participant contracted active

MRSA and

was switched

from NIG to ONG on medical advice. ITT analysis performed- missing blood naltrexone levels were classified as

0

 

spectrometry.

Unsupervised urine drug screening

Y, yes; N, no; U, unclear; MMT, methadone maintenance treatment; ITT, intention to treat analysis; ONG, oral naltrexone group; NIG , naltrexone implant group; ^ Same base cohort

 

Comparative studies

 

Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
Carreño

2003

 

Y

First 156 patients treated under the NIMROD program (ROD using a combination of antagonists, followed by naltrexone implant plus individual motivational orientation and CBT. NIG: 98% males, mean age 28.8 years, 75% had a partner, 83.3% stable work, 79.1% middle to technical education level. 80.58% used heroin, 6.08% used methadone, and 13.33% used heroin and methadone, 14.6% injected

Y

Outpatients with

criteria of opioid dependence (ICD-10), compliance with the centres admission criteria, no concomitant disease, contraindications or conditions making follow-up difficult, willing to consent and started program between

September 1998 and

October 2000

 

N

Pre treatment, the NIG scored higher than the oral group on occupation/ support (p<0.05), no adjustments reported

U

U if included in analysis Y Reported: 101/156 (65%) completed the program. Of these, only 56% returned for 6 month follow up and 21% for 12 month follow up

U U Y

EuropASI and retention index (assessed objectively).

EuroASI assesses psychiatric status, family/social relationships, legal status, drug use, alcohol use, occupation/ support and medical status

N
Colquho

un 2005

 

Y

Opioid dependent outpatients. NIG: 25 (61%) males, 16 (39%) females.

Mean age 26.2 years, mean years using opiates 7.2 (sd 5.0), mean years of education 10.6 years. A large number of participants in both groups were moderately to highly depressed

Y

Clients had to be motivated to be opiate free, have suitable social support and no psychiatric diagnoses

or medical issues that

would make

detoxification dangerous

N

No significant difference (p<0.05) between groups in age, years using opioids, years of education, daily opioid dose, psychiatric symptoms, and depression. Patients self selected treatment group

Y N

Number of counselling sessions was not significantly different between the

groups

 

N N

Self-esteem and quality of relationships using a 10 point Likert scale. Both administered via telephone survey of participants and their support person

U
Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
Grusser

2006

 

N

Small sample. NIG: 3 females and 14 males, mean age 34.76 years (sd 9.63). The mean duration of heroin use across all four groups was 7.02 years

(sd 5.82)

N N

In part. Age, duration of heroin use and gender were considered but patients in groups one and two chose their treatment

N

Reported but not included in analysis at 6 and

12 weeks

U U U

Unclear how relapse rates were assessed. Drug dependence: DSM-IV-TR

Craving and current mood: Questionnaire on

Differentiated

Assessment of Addiction Depression:

General

Depression Scale

 

U
Hulse 2005a^ Y

Opioid dependent persons who had not previously been treated with a naltrexone implant. 218 (60%) male, 143 (40%) female. Mean duration of heroin use 5.7 years (sd 5.3). 174 (48%) prior oral naltrexone maintenance, with 17% having entered into oral naltrexone maintenance >3 times. Males were significantly older than the females [mean 28.5, sd 7.2 vs mean 26.6 sd 7.9; t 2.4, p

<0.017]

Y

Opioid users treated for the first time with a naltrexone implant at a not for profit

community based clinic in Perth between Jan 2001 and Dec 2002. 361 (94%) of 384 possible participants considered eligible,

i.e. could be followed up via the WALD. Exclusion: no consent given (5), implants removed in the first week (3), and incarceration prior to implant (15)

N

In part. Age, age of first heroin use, duration of heroin use, gender, previous use of oral naltrexone considered. Mentioned confounder of age and sex in relation to mortality

Y

1 participant died and 1 had an early removal, both were included in the analysis. 326 (90%) were in the WALD, 257 (71%) were in the EDIS and 335 (93%) were in both

U Y

Prospectively collected data: assessor blind to treatment

allocation at outcome

assessment

 

Y

Prospectively collected hospital admission and mortality data

from the WALD and the EDIS. Overdoses were identified using

ICD10 codes

N/A

due to record linkage design

Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
Ngo

2008a^

Y

As per Hulse 2005a

Y

As per Hulse 2005a

Y

Potential confounders such as age, gender, preexisting history of mental illness and length of heroin use were

considered

 

Y

1 participant died

U Y

Prospectively collected data: assessor blind to treatment allocation at outcome

assessment

 

Y

Prospectively collected hospital admission and mortality data from the WALD  cases identified using ICD-9-CM diagnosis codes

Y

2 cases with missing data were removed from analysis

Ngo

2008b^

Y

Base cohort as per Hulse

2005a

 

NIG: 314; 129 (41%) females

[mean age 27.1 years sd 8.1],

185 (59%) male [mean age

29.0 years sd 7.5]

 

MMT: 522; 198 (37.9%) females [mean age 30.4 years sd 9.1 years],

324 (62.1%) males [mean age

32.1 years sd 8.9 years]

Y

As per Hulse 2005a

 

MMT cohort

ineligible if taking methadone for pain management, treatment crossover, or incarcerated  

 

N

Age and sex and their modifying

interactions were examined but other factors, e.g. level of motivation, situational influences, socioeconomic background, and pre existing

illness were not

 

N

82.6% MMT, 92.2% NIG lost to follow up, not included in analysis

U The MMT

cohort was only included as a reference group, so between group comparisons must be viewed with great caution

Y

Hospital morbidity and mortality collected prospectively and independent of research

team

 

Y

Prospectively collected hospital morbidity data

from the

WALD.

Physician coded mental health hospital admission using ICD 9 or 10. Inconsistencies in coding between different ICD codes are inevitable

N/A

due to record linkage design

Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
Ngo 2007^ Y

As per Hulse 2005a

Y

As per Hulse 2005a

Y

Investigated influence of gender, age, prior history of hospital admission, and duration of heroin use. Pre treatment truncated to match post

treatment

 

U

Not reported. Likely to be as

per Hulse 2005a

 

U Y

Hospital morbidity and mortality collected prospectively and independent of research

team

 

 

Y

Physician coded mental health hospital admission using ICD-9 and 10, but use of diagnostic codes from different versions means coding inconsistencies

are inevitable

 

N/A

due to record linkage design

Reece

2007

 

Y

Patients presenting to a low cost private naltrexone clinic in Brisbane.

NIG:  69 (68%) male, 33(32%) female. mean age 26.1 years, mean duration of opiate use 5.5 years. 9% on social security alone, 64% in socioeconomic class III-V and

27% in socioeconomic class III. 25% had ‘weak’ social support, 51% ‘average’ and

25% ‘strong’

N

No information on types of participants eligible for programs offered by the clinic

N

Matched for age, sex, duration of opioid use and dose used. Participants selected treatment. Financial status and social support better in the NIG, differences not

controlled for

 

Y

Overall followup was 82%. Urine drug screen followup was 76%. ITT analysis performed

U U N

Primary outcome was ‘opiate free success’, defined as a composite measure including self report, having a urine drug screen negative for opiates or having a carer report satisfactory progress

U
Tait

2008a^

Y

Base cohort as per Hulse 2005a, but restricted to treatment resistant group. Mean history of heroin dependency 7.5 years (sd 4.9),

Y

As per Hulse 2005a had to have received oral naltrexone 12 months prior to their implant, leaving

U Y

As per Hulse

2005a  1 participant died

Y 24 (19%) received an agonist therapy post implant, Y

Mortality data collected prospectively and independent

Y

ICD-9 or 10

N/A

due to record linkage design

Study Sample representative of the population  Recruitment/

eligibility criteria reported

Consideration and strategies to control, possible confounding factors Losses to follow up reported and included in analysis Other interventions received differentially during followup Outcome

assessors blind to treatment allocation

Outcomes measured in a reliable way Missing data (group or time-point data) accounted for
which is higher than average. 59% male, 41% female. 45% had only ever received oral and implantable naltrexone, 34% reported use of agonist treatment prior to implant treatment

 

130/139 (93.5%) mainly MMT, dates are unavailable of research

team

 

Tait

2008b^

Y

Base cohort as per Hulse

2005a

NIG: 40% female [mean age

27.0 years sd 7.9], 60% male

[mean age 28.8 years; sd 7.4]

 

Y

As per Hulse 2005a

Y

Only for age and gender

Y

As per Hulse

2005a

U

Additional treatment episodes possible but not reported

Y

Mortality data collected prospectively and independent of research team

Y

WALD which compiles extensive health information for individuals

N/A

due to record linkage design

Waal 2006 Y

Actively seeking this type of treatment. Small sample, 11 male, 2 female [mean age 26.9 sd 4.9]. Mean history of heroin dependency 4.8 years (sd 3.3). Most had extensive experience

with several other psychoactive substances. Participants had received a range of previous treatments

Y

Participants were actively interested in naltrexone implants

U

 

Y

2 participants interviewed with retrospective data

U N/A

due to pre/ post test deign

Y

Liquid chromatography/ mass spectrometry operated in electrospray ionisation, Hopkins

symptom

checklist, Beck depression rating scale, side effects checklist, Likert scales

 

N

Compliance with monthly evaluation procedures varied. Only 7/13 participants accounted for on mental health and craving outcomes at week 12

Y, yes; N, no; U, unclear; MMT, methadone maintenance treatment; NIG naltrexone implant group; ^ Same base cohort; ROD, rapid opioid detoxification; WALD, West

Australian Health Services Research Linked Database

Studies with no control group

 

Study Was the sample representative of patients in the population as a whole? Were recruitment/ eligibility criteria reported? Were losses to follow-up reported and included in analysis? Were losses to follow up> 20%? Were outcomes measured in a reliable way? Were missing data (group or time-point data) accounted for?
Olsen 2004 Y

Small sample. 5 male, 5 female [median age 30.5 (23-29)]. Median heroin use of 7.5 years (4-15). All HCV positive. Under no obligation to abstain from drug use

 

Y

Participants had to be taking part in a professional treatment or counselling program

Y

 

N

3 participants were lost to follow up. Adverse effect results reported for all participants

U

 

N/A

No missing data

Foster 2003 Y

Opioid dependent outpatients, seeking treatment in private clinic.

Cohort one: 76% male, 24% female [mean age 29.8 years sd 5.57]. 51% unemployed, 33% current and 67% previous injectors, mean opiate habit

6.7 years

Cohort two: 83% male, 17% female [mean age 27.4 years]. Comprehensive demographic data not collected

 

N Y N N

Self report (semi structured interview be telephone) with corroboration by family members/carers where possible. Craving on a 10 point visual analogue scale. Blood concentrations measured in several patients not in these cohorts

N/A
Y, yes; N, No; U, unclear; HCV, hepatitis C virus

 

APPENDIX E: RESULTS OF STUDIES ON EFFECTIVENESS

 

1)     Opioid use during and after treatment

 

RCTs

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Kunøe

2009

 

Norway

56

(29/27)

Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant 2200mg Usual care, i.e. encouraged to contact relevant aftercare: outpatient counselling, Norwegian maintenance treatment program, re admission to detoxification, or residential treatment 6 months Between group comparison

a)                6 month follow up: heroin use significantly lower in NIG (mean 17.9 days) vs controls (mean 63.6 days) difference 45.6 p<0.05, F 7.0 [95% CI 14.1, 77.3] ASI 30 day variable, NIG mean 3.5 days of heroin use vs 11.4 days in controls, difference 8, p<0.05, F 5.8 [95% CI 1.8, 14], and ASI frequency scale, difference 0.73, p<0.05 [95% CI 0.11, 1.34]

b)                6 month follow up for all opioid use, NIG scored significantly lower

(mean 37 days) than controls (mean 97.1 days) difference 60.2 p<0.01,

F 8.1 [95% CI 20.9, 99.5] ASI 30 day variable, NIG mean 6.3 days of opioid use vs 17.4 days in controls, difference 9, p<0.01, F 5.4 [95% CI 1.6, 16.4], and ASI frequency scale, difference p<0.001 [95% CI 0.45, 1.7].

c)                Abstinence from all opioids for the whole period was not significant, 11/29 in the implant group vs 5/27 in controls.

d)                At the 6 month follow-up assessment opioid dependence as per DSM-IV diagnoses using MINI was significantly lower in the implant group (9/29) vs controls (18/27), OR= 0.225, p=0.015 [95% CI 0.07, 0.69]

Note: The results of hair analysis matched self-reported opioid use in patients available for testing (86%).

 

Hulse

2009

 

Australia

70 (35/34)

 

Opioid dependant outpatients who had completed preclinical screening, exclusion criteria detailed Go Medical naltrexone

implant (2.3g)

 

Oral naltrexone

tablets

(50mg/d)

All underwent detoxification and were implanted with active or placebo implant.  All were encouraged to attend weekly individual, group or family therapy 6 months

 

Between group comparison

At 6 month follow-up, heroin use was significantly lower in the implant group

[hazard ratio 4.49; 95%CI 1.85,10.90]

Return to regular heroin use occurred significantly earlier in the oral group (median 115 days SE 12.0) compared to the implant group (median 158 days SE 9.4) p=0.001.

12% and 6% of oral and NIG, respectively, returned to heroin 1 to 3 days per week.

0% and 14% of oral and NIG, respectively, returned to heroin 1 to 3 days per month. Self reported complete abstinence in 27% and 63% of the oral and implant groups respectively.

 

NHMRC Literature Review: Naltrexone Implants for Opioid Dependence                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                39

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Colquhou

n 2005

 

Australia

83

(41/42)

Opioid dependent outpatients Go Medical naltrexone

implant (1700mg

single or double) plus counselling and regular phone support

Oral naltrexone plus counselling and regular phone support 12 months (plus 8 months during

treatment)

 

Between group comparison

Abstinent participants were those who had used opioids only a few times since detoxification.

 

In patients who had remained abstinent from opioids, relapse in the NIG was higher than control at 6 months post treatment, 33/41 (80%) vs 23/42 (55%) respectively. This trend continued at 12 months post treatment, where rates were 25/41 (61%), vs 17/42 (40%) respectively. No ratio or CI reported.

 

In patients who had relapsed from opioids (including those non contactable) relapse in the NIG was lower than controls at 6 months post treatment, 8/41(20%) vs 19/42(45%) respectively. This trend continued at 12 months post treatment, where rates were 16/41 (39%), vs 25/42 (60%) respectively. No ratio or CI reported

 

Hulse

2005a^

 

Australia

 

361

 

Opioid dependent

outpatients

 

Go Medical naltrexone implant (2200mg) simultaneously

with ROD

 

6 months pre treatment 6 months

 

Within group comparison

For emergency department presentations and hospital admissions and when data from emergency department presentations and hospital admissions were combined, there were more opioid overdoses 6 months pre treatment vs 6 months post treatment but the significance was not reported. Refer to Ngo 2008b

 

Ngo

2008b^

 

Australia

 

836

(314/ 522)

Opioid dependent outpatients Go Medical naltrexone implant (2200mg) simultaneously with ROD MMT

(methadone syrup prescribed by medical practitioners and dispensed by pharmacists)

3.5 years (plus

6 months pre treatment)

 

Between group comparison

There was no significant change in risk for other opioid hospitalisations from 6 months pre treatment to 6 months post treatment for either cohort. In the MMT cohort there was no significant change in risk of other opioid hospitalisations pre treatment to 3.5 years post treatment whereas for the implant cohort it was significantly reduced, 181 events vs 332 events respectively,

OR 0.64 [95% CI 0.46, 0.89], α = 0.05

There was a significant decline in hospital admission rates from opioid overdose from 6 months pre treatment to 3.5 years post treatment in the NIG,

RR 0.29 [95% CI 0.15, 0.55], α = 0.05 whereas the change was not significant in the MMT cohort.

There was a significant increase in hospital admission rates for other opioid conditions from 6 months pre treatment to 3.5 years post treatment in the MMT cohort, RR 1.35 [95% CI 1.14-1.61], α = 0.05 vs a significant decline in the NIG RR 0.55 [95% CI 0.46-0.65], α = 0.05. In the NIG female patients 30 years and older incurred fewer other opioid hospital admissions than their younger counterparts, RR 0.32 [95% CI 0.16-0.65], α = 0.05

 

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Reece

2007

 

Australia

376

(102/ 113/

161)

Opioid dependent outpatients Go Medical and

Wedgewood

naltrexone implants (1.1g and 1g respectively) Many had multiple implants

Revia naltrexone

tablets

(50mg/d)

a)                concurrent control

b)                historical control (12 months prior to NIG)

12 months

 

Within group comparison

Within the NIG social support, work prior to treatment and the presence of cannabis in the urine were significantly related to opiate free success.

Tait

2008a^

 

Australia

130 Opioid dependent

outpatients who had received oral naltrexone prior to implant

Go Medical naltrexone implant (2200mg) 6 months pre and post: first heroin use, oral naltrexone treatment, implant 6 months Within group comparison

There was a significant decline in opioid overdoses from 6 months before oral naltrexone to 6 months post implant treatment (Wilcoxon Z= 2.3, p=0.02).

 

Opioid overdoses increased from 0 in both the 6 months before and after first heroin use, to 7 in the both the 6 months before and after oral naltrexone to 2 in the 6 months before implant and 0 in the 6 months post implant (Friedman 2 15.2 (5), p =0.01).

MMT, methadone maintenance treatment; NIG, naltrexone implant group; ROD, rapid opioid detoxification; ^ same base cohort

 

2) Treatment adherence

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent

outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg)

Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare: outpatient

counselling, Norwegian maintenance treatment program, re admission to detoxification, or residential treatment

6 months Between group comparison

At 6 months follow up there was no significant difference in outpatient treatment attendance between the implant group and controls, p>0.05.

 

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Grusser

2006

 

Germany

68

(17/17 )

Opioid dependent patients UROD with

oral naltrexone, followed by one week of oral naltrexone treatment (50mg/d), followed by naltrexone implant (1000mg)

2)                Detoxified and treated with oral

Levomethadon

e

3)                Actively consuming

opioid addicts  4) Healthy volunteers

12 weeks compared with data collected 4 days after detoxificat ion, and 6 weeks into

treatment

 

Within group comparison

a)  6 weeks after treatment:

There was one relapse in the NIG.

Only 4 participants remained in group 2; 10 had relapsed and 3 had switched to long-term Levomethadone without any intention of becoming drug free. No significance reported.

 

b)  12 weeks after treatment

2 participants in the NIG had relapsed. None of the remaining 4 subjects in group 2 could be located. No significance reported.

 

Reece

2007

 

Australia

 

376 (102/

113/ 161)

Opioid dependent outpatients Go Medical and

Wedgewood

naltrexone implants (1.1g and 1g respectively), many had multiple implants

Revia naltrexone

tablets

(50mg/d)

a)                concurrent control

b)                historical control (12 months prior to NIG and

above)

 

12 months

 

Between group comparison

The proportion of patients in each group whose urine tested positive for naltrexone were 70%, 52% and 39% respectively.

NIG, naltrexone implant group; UROD, ultrarapid opioid detoxification; ^ same base cohort

 

3)  Retention in treatment: time to drop out

 

Comparative studies

Study/ Country N (n/group ) Population Intervention Comparator Follow-up Main Findings
Carreño

2003

 

Spain

440

(156/ 284)

Opioid dependent outpatients UROD

combination of antagonists

followed by Wedgewood

naltrexone implant

(1000mg) and

CBT

 

Classic detoxification followed by maintenance with oral naltrexone and

CBT

12 months Between group comparison

At 6 and 12 months post treatment the treatment retention index was significantly greater in the implant group (80% at 6 months, 65% at 12 months) than the oral group (42% at 6 months, 17% at 12 months), p<0.05. No CI reported.

Colquhou

n 2005

 

Australia

 

83

(41/ 42)

Opioid dependent outpatients Go Medical naltrexone

implant (1700mg

single or double) plus counselling and regular phone support

Oral naltrexone plus counselling and regular phone support 12 months (plus 8 months during

treatment)

 

Abstinent participants were those who had used opioids only a few times since detoxification.

 

Within group comparisons

In the NIG the time compliant to naltrexone was greater in the abstinent group than those who had relapsed at 6 and 12 months, p<0.05. NIG participants abstinent at 6 and 12 months estimated the implant was effective for 6 months. Those NIG participants who relapsed estimated the implant was effective for 4 months, p<0.05.

 

Between group comparisons

In patients who had remained abstinent from opioids, time spent in treatment was higher in the NIG than the ONG at 6 months post treatment, mean 176.6 days (sd 68.1) vs mean 120 days (sd 104.8) respectively. No p value or CI reported. This trend remained at 12 months post treatment [mean 187.3 days (sd 69.1) vs 123.21 days (sd 105.5) respectively]. No p value or CI reported.

 

In patients who had relapsed to opioid use, time spent in treatment was higher in the

NIG than the ONG at 6 months post treatment, [mean 112.5 days (sd 50) vs mean 19.7 days (sd 31.7) respectively]. No p value or CI reported. This trend remained at 12 months post treatment [mean 120 days (sd 45.6) vs 30.1 days (sd 54.25) respectively]. No p value or CI reported.

 

NIG, naltrexone implant group; ONG, oral naltrexone group; UROD, ultrarapid opioid detoxification

5)  Use of drugs other than opioids during and after treatment

RCTs

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Between group comparison

At 6 month follow up there was no significant difference in self reported use of alcohol or non opioid drugs between the NIG and controls, p>0.05.

 

Hulse

2009

 

Australia

70 (35/34)

 

Opioid dependant outpatients who had completed preclinical screening,  exclusion criteria detailed

 

Go Medical naltrexone implant

(2.3g)

 

Oral naltrexone

tablets

(50mg/d)

All underwent detoxification and were implanted with an active or placebo implant. All participants were encouraged to attend weekly individual, group or family therapy.

 

6 months Between group comparison

Overall use of non opioid drugs was similar between groups [hazard ratio 0.58 95% CI 0.32, 1.05]

 

Most frequently reported category of non opioid drug was cannabis, followed by stimulants.

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Hulse

2005a^

 

Australia

361

 

Opioid dependent

outpatients

 

Go Medical naltrexone implant (2200mg) simultaneously with ROD 6 months pre treatment 6 months

 

Within group comparison

There were more sedative overdoses in the 6 months post treatment than the 6 months pre treatment as measured by emergency department presentations and hospital admissions, significance not reported. Refer to Ngo 2008b

 

The number of other drug overdoses was the same in the 6 months pre treatment as the 6 months post treatment, as measured by emergency department presentations.

There were more other drug overdoses in the 6 months post treatment than the 6 months pre treatment as measured by hospital admissions, significance not reported.

 

When hospital admissions and emergency department presentations were combined, there were less sedative overdoses in the 6 months pre treatment (8, 1.9%) than the 6 months post treatment (16, 4.4%) p <0.01. Nine overdoses were in the first 10 days after treatment and if excluded the trend is neutral or reversed. Other drug overdoses increased from the 6 months pre treatment (2, 0.6%) to post treatment (5, 1.4%), significance not reported.

 

Ngo

2008b^

 

Australia

836  (314/

522)

Opioid dependent outpatients Go Medical naltrexone implant (2200mg) simultaneously with ROD MMT

(methadone syrup prescribed by medical practitioners and dispensed by pharmacists)

3.5 years (plus 6 months pre

treatment)

 

Between group comparison

Comparing non opioid overdose admissions from 6 months pre treatment to

6 months post treatment, no significant change in either cohort for patients aged

25 years, significantly increased risk in patients aged 35 years. MMT: OR 5.03 [95 % CI 1.18, 21.54], α = 0.05

NIG: OR 16.31 [95% CI 3.07, 86.53], α = 0.05

There was no significant change in risk of non opioid oversdose admissions from 6 months pre treatment to 3.5 years post treatment for either cohort.

 

Within group comparison

MMT: no significant change in risk of other non opioid admissions 6 months pre treatment to 6 months or 3.5 years post treatment

NIG: significantly increased risk at both follow-up periods,

RR 1.33 [95% CI 1.01, 1.73], α = 0.05. 12  vs 31 events at 6 months OR 2.54 [95% CI 1.19, 5.43], α = 0.05 and 105 vs 138 events at 3.5 years, OR 1.52 [95% CI 1.04, 2.23], α = 0.05

In the NIG reduced risk for each one year age increment, at 3.5 years the OR was 0.96 [95% CI 0.94, 0.98].

 

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Reece

2007

 

Australia

376

(102/ 113/

161)

Opioid dependent outpatients Go Medical and

Wedgewood

naltrexone implants (1.1g and 1g respectively), many had multiple implants

Revia naltrexone

tablets

(50mg/d)

a)   concurrent control

b)   historical control (12

months prior to NIG and

above)

 

12 months

 

Between group comparison

Urinary analysis was performed to test use of amphetamines, morphine,

Tetrahydrocannabinol (THC), cocaine and methadone. The rate of obtaining a urine drug screen for each of the groups was 98%, 70% and 76% respectively.

Amphetamine use was significantly greater in the NIG than either the concurrent or historical ONG or compared to both controls combined,  Chi Square 7.13, df 1, OR 2.78 [95% CI 1.22, 6.42], p=0.0075 Chi Square 11.14, df 1, OR 3.28 [95% CI 1.52, 7.18], p<0.001

Chi Square 13.83, df 1, OR 3.05 [95% CI 1.60, 5.83], p=0.0002 respectively

 

MMT, methadone maintenance treatment; NIG, naltrexone implant group; ONG, oral naltrexone group; ROD, rapid opioid detoxification; ^ same base cohort

 

6)  Criminal activity and incarceration

 

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Between group comparison

At 6 months follow up there was no significant difference in self reported criminal activity between the implant group and controls, p>0.05.

 

7)            Quality of life

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Between group comparison

a)                Life satisfaction: at 6 month follow up NIG scored higher than controls on the

Temporal Satisfaction With Life Scale, mean difference 5.4, p<0.05, F 5.3 [95% CI

0.68, 10.1]

b)                Recommend treatment to a friend: at 6 month follow up NIG scored higher than controls, mean 85 vs 56.5 respectively, on a visual analogue scale of treatment satisfaction, mean difference 28.5, p<0.05, F 9.1 [95% CI 10.8, 46.2].

c)                Satisfaction with treatment allocation: at 6 month follow up NIG scored higher than controls, mean 78 vs 36 respectively, mean difference 42, p<0.01, F 25 [95% CI 25.9, 58.5]

 

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Carreño

2003

 

Spain

440

(156/ 284)

Opioid dependent outpatients ROD using a combination of antagonists followed by long term maintenance Wedgewood

naltrexone implant

(1000mg) and

CBT

 

Classic detoxification followed by maintenance with oral

naltrexone and

CBT

12 months Between group comparison

No significant differences were found between the two groups at either follow-up, except at 6 months post treatment the NIG scored better than the ONG on family/social relationships, p= 0.05, no CI reported

 

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Colquhoun

2005

 

Australia

83

(41/ 42)

Opioid dependent outpatients Go Medical naltrexone implant

(1700mg  single or double) plus counselling and regular phone support

Oral naltrexone plus counselling and regular phone support 12 months (plus 8 months during

treatment)

 

Abstinent participants were those who had used opioids only a few times since detoxification.

 

Within group comparisons: in participants who were successful in abstaining from opioid use there was a significant increase in self esteem and quality relationships between pre treatment and 6 and 12 months post treatment, p<0.01, in both groups. No CI reported.

Between group comparisons: in participants successful in abstaining from opioid use, the differences between the NIG and the ONG on measures of self esteem and relationship quality at pre treatment were not significant, p=0.86 and 0.81 respectively. In this same group at 6 months post treatment, the NIG had lower self esteem ratings than the ONG (mean difference 1.2), p=0.018. No CI reported. The ONG also tended to score better on relationships at 6 months post treatment compared to the NIG, but the difference was not significant, p=0.055.

Among abstinent participants, there was no significant difference in self esteem or relationship quality at 12 months post treatment, p>0.05

 

Reece

2007

 

Australia

376

(102/ 113/

161)

Opioid dependent outpatients Go Medical and Wedgewood

naltrexone implants (1g and 1.1g respectively), many had multiple

implants

 

Revia naltrexone

tablets, (50mg/d)

c)   concurrent control

d)   historical control (12 months prior to NIG and

above)

 

1 month

for this aspect of the study (plus 1 month before

ROD)

 

Within group comparison

Within the NIG and concurrent ONG work status improved from before to after treatment,

NIG: 18% pre treatment vs 50% after treatment, weighted OR 4.67 [95% CI 2.35, 9.33], p<0.001

Concurrent ONG: 25% pre treatment vs 48% post treatment

weighted OR 2.78 [95% CI 1.52, 5.09], p<0.001. The improvement in the historic ONG was not significant, p=0.178

NIG, naltrexone implant group; ONG, oral naltrexone group; ROD, rapid opioid detoxification; CBT, cognitive behavioural psychotherapy; ^ same base cohort

8)  Mental health

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Between group comparisons

At 6 month follow up there was no significant difference in self reported depression (Beck Depression Inventory and subscale of the 25-item Hopkins Symptom Checklist) between the implant group and controls, p>0.05.

 

 

 

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Ngo

2008a^

 

Australia

359 Opioid dependant outpatients Go Medical naltrexone implant (2200mg) simultaneously with ROD Pre treatment period truncated to match post treatment follow-up 1.78 years posttreatment Within group comparison

Counts of patients and hospital events was categorised into non-substance related, mood related, substance related or all-cause.

 

1. Comparison of pre- versus post-treatment:

a) risk of patient hospitalisation for a mental condition 

The number of cases for patients after naltrexone treatment was consistently lower than pre-treatment. However there was not a statistically significant difference. b) rate of mental health related hospital admissions

There was a reduction in mental health related hospital events post naltrexone treatment. There was a significant reduction in admissions for non-substance mental disorders IRR .630 [95% CI .478, .841 p=.0017]. This was mostly observed in male patients (p=.0002) compared to females (p=.5363). There was a significant reduction in admissions for substance-related mental disorders IRR .673 [95% CI .558, .811 p<.0001] and All-cause mental disorders IRR .641 [95%

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
CI .536, .766 p<.0001]

 

2. Evaluating contributing factors to risk of patient hospitalised due to mental illness following treatment:

Higher risk:

-  of hospitalisation for NSRMD with history of hospitalisation for NSRMD OR

4.78 [95% CI 2.53, 9.01 p<.0001]. Females at higher risk OR 2.39 [95% CI

1.25, 4.58 p.0085].

-  >5 years of heroin dependence OR 3.82 [95% CI 1.08, 13.5 p.0378].

-  of hospitalisation for substance related mental disorder with history of hospitalisation for substance related mental disorder OR 1.82 [95% CI 1.10, 3.04 p.0208]. Females at higher risk OR 1.95 [95% CI 1.21, 3.15 p.0063].

 

3. Length of stay

a)        median per episode varied with different time points in the younger cohort, effect not seen in older patients.

b)        total per person year increased by 1.6 days post treatment, which is similar to the ‘during use’ period.

 

Ngo

2007^

 

Australia

359 Opioid dependent outpatients Go Medical naltrexone implant (2200mg) Pre treatment period truncated to match post treatment follow-up 6 months Within group comparison

The number of hospital admissions significantly decreased pre to post treatment for NSRMD [RR 0.630 95% CI 0.472, 0.841, p = 0.0017], substance related mental disorder [RR 0.673 95% CI 0.558, 0.811, p<0.0001], and all cause mental disorder [RR 0.641 95% CI 0.536, 0.766, p<0.0001]. The number of hospital admissions decreased pre to post treatment for mood disorders, but after adjusting for confounding effects the trend was not significant (p= 0.4232), declining mainly among males [RR 0.463 95% CI 0.220- 0.974] and participants treated at a later age (26 years or older), [RR 0.356 95% CI 0.190-0.668].

There was no significant difference in risk of hospitalisation pre vs post treatment for NSRMD [OR 1.309 95% CI 0.877, 1.954], mood disorders [OR 1.283 95%

CI 0.689, 2.390], substance related mental disorders [OR 1.214 95%

CI 0.886, 1.664], or all cause mental disorders [OR 1.272 95% CI 0.931, 1.738]. Higher risk of non substance related hospital event post treatment with prior hospitalisation for NSRMD OR 4.78 [95% CI 2.53, 9.01]. Females were at higher risk than males, OR 2.39 [95% CI 1.25- 4.58].

 

Participants with >5 years of heroin use pre treatment were at higher risk from hospitalisation due to a mood disorder post treatment compared to those who had used for a shorter period, OR 3.82 [95% CI 1.08, 13.5].

Participants with a history of substance-use hospital diagnoses were at higher risk

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
from hospitalisation due to substance use disorder post treatment OR 1.82 [95% CI 1.10, 3.04]. Females were at higher risk than males, OR 1.95 [95% CI 1.21, 3.15]. Older age at treatment was a protective factor for not developing a drug-related problem post treatment OR 0.949 (0.949- 0.958 for every one year age increment).

Females were at greater risk of all cause mental disorders than males, OR 1.87 [95% CI 1.16, 3.01]. Older age at treatment was a protective factor, OR 0.951 (0.942 0.960 for every one year age increment).

Length of hospital stay due to mental disorders was not significantly different from pre to post treatment, p=04120.

 

Tait

2008a^

 

Australia

130 Opioid dependent

outpatients who had received oral naltrexone prior to implant

Go Medical naltrexone implant (2200mg) 6 months for each: pre and post first heroin use,  Pre and post oral naltrexone, pre and post implant 11.3 months Within group comparison

General trend for an increase in mental health admissions with an opioid related diagnoses in the 6 months post first heroin use, pre and post oral naltrexone periods, with a decline in the 6 months pre and post implant (Friedman65.2 (5), p <0.001). There was a significant decline in mental health admissions with an opioid related diagnoses from 6 months before oral naltrexone to 6 months post implant treatment (Wilcoxon Z = 2.3, p<0.0001).

The trend in mental health admissions with a drug related diagnosis other than opioids was not significant over the 6 study periods (Friedman9.9 (5) p =0.082). In the 6 months pre implant there were two non opioid mental health admissions, 1 with stimulants and 1 with stimulants plus cannabis. In the 6 months post implant there 5 non opioid mental health admissions, 2 with unspecified multidrug use, and 1 each for sedatives, stimulants and cannabis use.

The trend in mental health admissions excluding any drug related diagnoses was not significant over the 6 study periods (Friedman3.6 (5) p =0.616).

 

Waal

2006

 

Australia

13 Opioid dependent outpatients Go Medical naltrexone implant (1800mg

single, double or single + double)

1 week pre treatment 11.3 months Within group comparison

Anxiety/stress levels remained relatively stable pre and post treatment, [ mean 1.9 (SD 0.69) pre treatment vs 1.3 (SD 0.32) to 1.7 (SD 0.58) post treatment] P values not reported.

 

Depression levels generally decreased from 1 week pre treatment to 8 weeks post treatment, [ mean 16.8 (SD 13.8) pre treatment vs 7.6 (SD 7.2) to 12.0 (SD 12.1) in the 8 weeks post treatment]. Depression scores increased at the end of the 12 weeks post treatment[ mean 14.7 (SD 10.4)] P values not reported.

 

NSEMD, non substance related mental disorder; ^same base cohort

9)  Duration of achieved therapeutic naltrexone blood levels

RCTS

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential

treatment

 

6 months Within group comparison

Plasma levels of naltrexone stayed above 1ng/ml for 6 months and above 2ng/ml for about 5 months. Levels of 6-β-naltrexone had a similar distribution.

 

Note: only 14 of 29 (48%) participants presented for testing of plasma levels.

 

Hulse

2009

 

Australia

70 (35/34)

 

Opioid dependant outpatients who had completed preclinical screening, exclusion criteria detailed Go Medical naltrexone implant

(2.3g)

 

Oral naltrexone

tablets

(50mg/d)

All underwent detoxification and were implanted with an active or placebo implant.  All participants were encouraged to attend weekly individual, group or family therapy.

 

6 months Between group comparisons

Significantly more of the ONG had blood naltrexone levels <2ng/mL at months 1 (ONG 72% NIG 6%) and 2 (ONG 81% NIG 48%) and <1ng/mL at months 1 through 4 (month 1: ONG 59% NIG 0%; month 2: ONG 74% NIG 18%; month 3: ONG 80% NIG 53%; month 4: ONG 97% NIG 53%)

 

Blood naltrexone concentration estimates Men (mean Body Mass Index 25.5, age 33):

≥ 2ng/ml 56 days [95% CI 39, 73]

≥ 1ng/ml 110 days [95% CI 83, 119]

 

Women (mean BMI 25.1, age 28):

≥ 2ng/ml 43 days [95% CI 16, 79]

≥ 1ng/ml 124 days [95% CI 88, 175]

 

 

Comparative study

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Waal

2006

 

Australia

13 Opioid dependent outpatients Go Medical naltrexone

implant (1800mg

single, double or single +

double)

 

1week pre treatment 11.3 months Within group comparison

For participants without measurable naltrexone before implantation, plasma levels were maintained above 1-2ng/ml between 1 and 3 months after a single implant and 3 and 5 months after a double implant. The 6-β-naltrexone concentrations were approximately 1-2.5 times higher compared to naltrexone during the whole implant period.

 

10)  Heroin craving

 

RCT

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent

outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg)

Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, application for

entry to the

Norwegian maintenance treatment program, re admission to detoxification, or residential treatment.

 

6 months Between group comparisons

At 6 month follow up the implant group scored lower than controls on a visual analogue scale of craving related thoughts,

mean difference 27, p<0.01, F 7.2 [95% CI 9.4, 44.2]

 

 

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Waal

2006

 

Australia

13 Opioid dependent outpatients Go Medical naltrexone

implant (1800mg

single, double or single +

double)

 

1 week pre treatment 11.3 months Within group comparison

On a scale of 1 (not at all) to 7 (very strong). craving dropped from pre treatment to

1-8 weeks post treatment [mean 2.6 (SD 1.0) pre treatment vs 2.4 (SD 1.6) to

2.7 (SD 1.4) post treatment]. Craving increased at week 12, when the possibility of relapse came closer [mean 3.1 (SD 2.1)], p values not reported.

 

RESULTS OF STUDIES ON SAFETY

Study/ Country N

(n/group)

Population Intervention Comparator Follow-up Main Findings
Kunøe

2009

 

Norway

56 (29/27) Opioid dependent outpatients who had completed abstinence oriented inpatient treatment and passed an oral naltrexone challenge (25mg) Go Medical naltrexone implant (2200mg) Usual care, i.e. encouraged to contact relevant aftercare such as out-patient counselling, Norwegian maintenance treatment program, re admission to detoxification, or residential treatment. 6 months a)                Site related: 2 participants reported wound opening with leakage of fluid, 3 had allergic reactions all of which were treated. 3 participants had implants removed (1 due to necrosis, 2 on request due to discomfort- site pain or diarrhoea).

b)                Possibly naltrexone related: 1 participant in the NIG overdosed using a combination of opioids, amphetamines and benzodiazepines, 4non fatal over doses were reported in control group.

c)                Mortality: 1 in NIG prior to implant due to overdose, 1 in control group due to overdose 3 months into study.

Hulse

2009

 

Australia

70 (35/34)

 

Opioid dependant outpatients who had completed preclinical screening, exclusion criteria detailed Go Medical naltrexone implant

(2.3g)

 

Oral naltrexone

tablets

(50mg/d)

All underwent detoxification and were implanted with an active or placebo implant.  All participants were encouraged to attend weekly individual, group or family therapy. 6 months a)                Site related: 1 participant had a severe adverse event related to the surgical procedure– wound hematoma proximal to the implant site.

b)                Possibly naltrexone related: 10 study related unexpected adverse events (4 oral, 6 implant) – typically diarrhoea or headaches.

29 study related expected adverse events (10 oral, 19 implant) – typically opiate withdrawal symptoms (diarrhoea, nausea, vomiting), exudation redness, pain at the surgical incision site or proximal to the implant in the first 14 days after surgery.  No opiate overdoses requiring hospital treatment or admission were reported during the 6 month follow up period.

NIG, naltrexone implant group

Comparative studies

Study/ Country N

(n/group)

Population Intervention Comparator Followup Main Findings
Carreño

2003

 

Spain

440

(156/ 284)

Opioid dependent outpatients ROD using a combination of antagonists followed by long term

maintenance Wedgewood

naltrexone implant

(1000mg) and

CBT

 

Classic detoxification followed by maintenance with oral

naltrexone and

CBT

12 months a) Site related: 7 (4.5%) had local allergic tissue reactions. 3 (1.9%) had local wound infection, but no implants had to be removed.

 

Hulse

2005a^

 

Australia

361 Opioid dependent

outpatients

 

Go Medical naltrexone implant (2200mg) simultaneously

with ROD

 

6 months pre treatment 6 months

 

a)                Site related: 3 implant removals in the first week at the patients request, 2 for psychological reasons and 1 due to infection at the wound site. Another removal at 169 days post implant due to an allergic reaction.

b)                Mortality: 1 death due to head trauma, coroners report not available

Ngo

2008b^

 

Australia

836

(314/ 522)

Opioid dependent outpatients Go Medical naltrexone implant (2200mg) simultaneously

with ROD

 

MMT

(methadone syrup prescribed by medical practitioners and dispensed by

pharmacists)

 

3.5 years (plus 6 months pre treatment)

 

a)                Site related: as per Hulse 2005a

b)                Mortality: 1 drug related death in the NIG vs 6 in the MMT group. In the MMT group 2 deaths were opioid related, 1 was a combined drug overdose, and 2 were non-opioid drug overdoses. The mortality in the NIG was from an opioid overdose in a participant previously treated with MMT.

Tait

2008b^

 

Australia

894

(341/ 553)

Opioid dependent outpatients Go Medical naltrexone implant (2200mg) MMT 3.5 to 5.5 years a)  Site related: as per Hulse 2005a

b)  Mortality: 6/341 (1.8%) versus 15/553 (2.7%).

The age standardized mortality rate ratio for naltrexone compared to methadone was 0.645 [95% CI 0.123, 1.17].

The mortality rates for the initial 14 day period’, ‘stable treatment (0-6 months) and overall’ were 0.0, 4.21 and 3.76 vs 94.47, 0.0 and 5.83/1,000 p-y.

In the NIG 2 deaths were drug related and 1 self inflicted. In the MMT group 5 cases were drug related and 2 were self-inflicted.

 

Waal

2006

 

Australia

13 Opioid dependent outpatients Go Medical naltrexone implant (1800mg

single, double or single + double)

1 week pre treatment 11.3 months a)                Site related: 2 local tissue reactions, 1 was mild and responded to antihistamines, the other (of which the patient had had 4 earlier implants) developed necrosis and required removal. At 12 weeks post treatment 9 could still feel their implant on palpitation and 6 felt some anxiety or discomfort by the lump.

b)                Possibly naltrexone related: Side effects (on a checklist including cephalgia, nausea, diarrhoea, muscle and articular pain or discomfort, anxiety and irritability) were recorded for all 13 participants. In the first week pre treatment, diarrhoea, muscle pain and irritability were the most frequency reported vs muscle pain, irritability and anxiety in the first week post treatment. The mean value of all side effect scores was in the range of 'not at all' to 'small' throughout the post implant observation period.

MMT, methadone maintenance treatment; NIG, naltrexone implant group; ROD, rapid opioid detoxification; CBT, cognitive behavioural psychotherapy; ^ same base cohort

Studies with no control group

Study/ Country N Population Intervention Followup Main Findings
Olsen

2004

 

Norway

10 Opioid dependent outpatients Wedgewood naltrexone implant (1000mg single, 3 or 4) 80 days a)                Site related: 3 implant removals (one each due to necrosis, local tissue reactions, on patient request due to mixed abstinence phenomena and desperation). First 2 removals were after repeated implantation.

b)                Possibly naltrexone related: the most prevalent side effects were irritability and dysphoria in the first week post implant (reported as 6/10 in Lobmaier 2008 not reported in study itself). Also reports of slight cephalagia, nausea and muscular discomfort (numbers not reported).

 

Foster

2003

 

England

 

 

101

(55/46)

Opioid dependent

patients

 

Wedgewood implant (1000mg single, double or single +

double)

1st cohort: ROD under general anesthesia followed by implant 2nd cohort: ROD (non-iv

sedation) followed by implant

 

12 weeks

 

a)                Site related: 15/101 (15%) had local tissue reactions. Most were mild and resolved without surgical treatment.

b)                Mortality: 2 deaths deemed to be unrelated to implant- 1 pulmonary embolism, 1 suicide in a participant with a history of depression.

ROD, rapid opioid detoxification