Slow-release oral morphine for opioid maintenance treatment: a systematic review

Abstract

This review article summarizes the results of all available clinical trials considering the use of slow-release oral morphine (SROM) for opioid maintenance treatment (OMT). All studies published up to October 2010 and assessing SROM for OMT in adult patients are included. Three independent reviewers assessed the selected articles using a standardized checklist. Study design, study length and number of subjects included were recorded. Data about retention rate (proportion of participants remaining under maintenance treatment at the end of the study), quality of life, withdrawal symptoms, craving, additional drug consumption, driving capacity and adverse events were collected. We identified 13 articles corresponding to nine clinical trials considering the use of SROM for OMT. Among them, only one was a randomized trial and one was a controlled not randomized trial. All other studies were uncontrolled. Retention rates were good (from 80.6 to 95%) with SROM maintenance, but similar retention rates were obtained with methadone. Most of the studies showed that quality of life, withdrawal symptoms, craving and additional drug consumption improved with SROM. However, there was no comparison with other maintenance drugs. As most of the studies assessing SROM efficacy were uncontrolled, there is no definite evidence that SROM is an effective alternative to methadone for OMT.

Objectives

United Nations Office on Drugs and Crime estimated that between 15 and 21 million people around the world were opiate users in 2007 [1]: 1.3 million in North America and 3.4–4 million in Europe. In Europe and Asia, opiate dependence remains the main drug problem as reflected in the number of treatments provided.

Opioid dependence represents a chronic relapsing condition that may require long-term maintenance treatment with oral synthetic opioid [2], as well as psychosocial interventions to re-establish social integration. Research has shown that heroin users in opioid maintenance treatment (OMT) buy less street heroin, consume fewer psychoactive drugs, have a reduced risk of HIV infection, need less income from crime to support their habit and are in less danger of fatal overdose, compared with an untreated sample [1, 3–5]. Unfortunately, only a small proportion of addicts receive OMT treatment for their opioid dependence. In the US and Europe, this is believed to be around 20% [1, 6].

Psychopharmacological drugs play an important role in rehabilitating patients and deserve special attention. The primary aim of pharmacological intervention is to reduce the number of relapses and to prolong the time between relapses. Maintaining quality of life and optimizing the patient's well-being are also of major importance and may need the use of more than one pharmacological agent. Evaluated drugs for OMT include methadone, buprenorphine and buprenorphine–naloxone combination [7]. The use of other opioid agonists as OMT has also been investigated with conflicting results [e.g. levo-α-acetylmethadol, prescribed heroin, codeine and slow-release oral morphine (SROM)].

The number of European countries where SROM is available as an alternative to standard methadone maintenance treatment is increasing: Austria (1998), Slovakia (2005), Slovenia (2005), Bulgaria (2006) and Luxembourg (2006) [8]. In France, where SROM can exceptionally be prescribed in the case of failure of methadone or buprenorphine previous maintenance treatment, the National Commission of Narcotics and Psychotropic Drugs failed to achieve a consensus decision about the use of SROM in OMT [9].

Therefore, the aim of this systematic review was to gather the results of all available clinical trials considering the use of SROM for OMT in order to summarize the scientific knowledge about this controversial subject.

Methods

All study types about SROM maintenance treatment were included (randomized controlled trial, controlled trial, prospective uncontrolled study and cross-sectional studies). There was no geographical or period restriction. We included articles in English, German or French published up to October 2010.

Different information sources were searched: PubMed, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials, Cochrane Drugs and Alcohol Group, Clinical trial.gov, National Institute on Drug Abuse, United Nations Office on Drugs and Crime, and European Monitoring Centre for Drugs and Drug Addiction.

Research equations used for searching each database are presented in Table 1. A Medline research equation was constructed using the following MeSH terms: publication type, type of maintenance drug (methadone, buprenorphine and morphine), substance related disorder and opiate alkaloids. Articles dealing with morphine analgesia were excluded by the research equation. Results were not limited to humans in order to avoid the risk of excluding relevant references without the assignment of ‘humans’ or ‘animals’ as an indexing term [10]. Other database searches were mainly performed using the keywords morphine and maintenance. Unpublished articles and congress presentations were not included. Finally, the references cited in selected articles were checked in order to include studies not found in any database.

Table 1.

Research strategy

Database	Research term
PubMed	1. ‘Randomized controlled trial’ [Publication Type] OR ‘controlled clinical trial’ [Publication Type] OR ‘comparative study’ [Publication Type]
	2. ‘Buprenorphine’ [Mesh] OR ‘methadone’ [Mesh] OR ‘morphine’ [Mesh:noexp] NOT ‘analgesia’ [Mesh])
	3. ‘Substance-related disorders’ [Mesh] AND ‘opiate alkaloids’ [Mesh]
	4. 1 AND 2 AND 3
Cochrane Database of Systematic Reviews	Morphine
Clinical trials.gov	(Morphine NOT analgesia NOT pain) AND Phase III IV
Cochrane Central Register of Controlled Trials	Morphine AND maintenance
Cochrane Drugs and Alcohol Group	Morphine – analgesia – pain
National Institute on Drug Abuse	Morphine AND maintenance
United Nations Office on Drugs and Crime	Morphine
European Monitoring Centre for Drugs and Drug Addiction	Morphine

The process of article selection was conducted by two independent reviewers (J. J. and M. L.-M.). Articles were first selected considering their title. Articles mentioning in their title ‘morphine’ or ‘maintenance treatment’ without further precision were selected. The abstracts of the remaining articles were then assessed. Experimental studies or studies not dealing with morphine were then excluded. The full text of remaining articles was assessed, and only articles about SROM maintenance treatment in nonpregnant addict patients were included in the analysis.

Three independent reviewers (J. J., A. G. and P. S.) assessed the selected articles using a standardized checklist. Article methodological quality was assessed using the CONSORT [11] or STROBE [12] checklist depending on the study design. Data about study design, duration and geographical location were collected. Participant characteristics (mean age, percentage of men and previous maintenance treatment), type of intervention and comparison group studied were recorded. Data about retention rate (proportion of participants remaining under maintenance treatment at the end of the study), quality of life, withdrawal symptoms, craving, additional drug consumption, driving capacity and adverse events were collected.

Risk of bias in individual studies was assessed by collecting data about study design and participant characteristics and by completing CONSORT and STROBE checklists. Study type and sample size were mentioned in data synthesis.

Because of methodological heterogeneity and the very low number of randomized controlled studies available, results were not combined, leading us to perform a qualitative systematic review. This review was conducted following the recommendations of the PRISMA statement (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) [13].

Results

A total of 1088 studies were identified through database searching. Figure 1 details the number of studies screened, assessed for eligibility and included in the review. Finally, 13 articles were included in our analysis [14–26].

Figure 1.

Flow diagram of number of studies screened, assessed for eligibility and included. SROM, slow-release oral morphine

Designs of included studies are summarized in Table 2. In fact, only nine different studies corresponded to the 13 selected articles; it appeared that different analyses were performed using the same data set (same study size, participants and intervention) for article numbers 2 and 6 [15, 19]; 3, 8 and 9 [16, 21, 22]; and 5 and 7 [18, 20]. Of these nine studies, only one was a randomized controlled trial, leading to two different articles [15, 19].

Table 2.

Articles included: summary of study design

Study design	Number of study	Article	Notes
Randomized controlled trial	1	2. Winklbaur et al. (2008) [15]	Crossover
		6. Eder (2005) [19]	Same data set for articles 2 and 6
Controlled trial	1	3. Mitchell et al. (2006) [16]	Crossover
		8. Mitchell et al. (2004) [21]
		9. Mitchell et al. (2003) [22]	Same data set for articles 3, 8 and 9
Prospective uncontrolled study	5	1. Kastelic et al. (2008) [14]	–
		12. Vasilev et al. (2006) [25]
		11. Kraigher et al. (2005) [24]
		10. Kraigher et al. (2002) [23]
		13. Rao et al. (2005) [26]
Cross-sectional study	2	4. Giacomuzzi et al. (2006) [17]	Same data set for articles 5 and 7
		5. Giacomuzzi et al. (2005a) [18]
		7. Giacomuzzi et al. (2005b) [20]

Synthesis of the main results is presented in Table 3. Retention rate was available in six studies of the nine included. It varied from 95% at 6 months [25] to 80.6% at 4 weeks [14], and was similar to the retention rate observed with methadone in the sole available comparative study (84.4% at 7 weeks with SROM vs. 90.6% with methadone) [19].

Table 3.

Synthesis of results

Outcome/article	n	Study type /duration	Effects estimated – 95% CI†
Retention rate
6. Eder et al. (2005) [19]	64	RCT/7 weeks	SROM: 84.4% (67.2%–94.7%)/methadone: 90.6% (75%–98%)
11. Kraigher et al. (2005) [24]	110	PUS/3 weeks	SROM: 93.6% (87.3%–97.4%)
1. Kastelic et al. (2008) [14]	67	PUS/4 weeks	SROM: 80.6% (69.1%–89.2%)
10. Kraigher et al. (2002) [23]	67	PUS/3 weeks	SROM: 94% (85.4%–98.4%)
13. Rao (2005) [26]	74	PUS/4 weeks	SROM: 44.6% (33%–56.6%)
12. Vasilev et al. (2006) [25]	20	PUS/6 months	SROM: 95% (75.1%–99.9%)
Quality of life
2. Winklbaur et al. (2008) [15]	64	RCT/14 weeks	Berlin Quality of Life Profile (BQLP): no statistically significant difference in any quality of life domain
8. Mitchell et al. (2004) [21]	18	CT/6 weeks	SROM was associated with higher SF-36 Social Functioning scores than methadone (mean score of 78 ± 18 vs. 58 ± 25)
1. Kastelic et al. (2008) [14]	67	PUS/4 weeks	Symptom CheckList 27 (SCL-27): enhanced well-being
12. Vasilev et al. (2006) [25]	20	PUS/6 months	Improvement of patient well-being
4. Giacomuzzi et al. (2006) [17]	240	CSS	Berlin Quality of Life Profile, overall satisfaction: methadone 5.3 ± 1.5 buprenorphine 4.9 ± 1.4 SROM 4.1 ± 1.7 no maintenance treatment 3.5 ± 1.6
Withdrawal symptoms
6. Eder et al. (2005) [19]	64	RCT/7 weeks	Significant decrease after week 1, withdrawal recurring during the crossover phase in both groups
8. Mitchell et al. (2004) [21]	18	CT/6 weeks	Withdrawal severity during the five first days of SROM maintenance was greater than during the resumption of methadone maintenance
11. Kraigher et al. (2005) [24]	110	PUS/3 weeks	Withdrawal symptoms decreased (mean Wang scale value of 12.1 at day 1 compared with 0.3 at day 20)
1. Kastelic et al. (2008) [14]	67	PUS/4 weeks	Decrease in the Short Opiate Withdrawal Scale (SOWS) value (in subgroup A 4.1 ± 3.7 at day 0 and 1.6 ± 3.0 at day 28, in subgroup B 9.0 ± 6.7 at day 0 and 2.9 ± 3.5 at day 28)
10. Kraigher et al. (2002) [23]	67	PUS/3 weeks	Decreased with SROM maintenance (mean Wang scale value of 11.2 at day 1 compared with 1.1 at day 20)
12. Vasilev et al. (2006) [25]	20	PUS/6 months	SOWS scores reduced markedly in the first 2 weeks and remained low until week 24 (mean value of 16.9 ± 6.6 at baseline vs. 6.6 ± 5.9 at week 24)
Craving
6. Eder et al. (2005) [19]	64	RCT/7 weeks	Craving for heroin, cocaine and alcohol decreased and remained low in both groups
8. Mitchell et al. (2004) [21]	18	CT/6 weeks	SROM was associated with reduced heroin craving compared with methadone [Visual Analogue Scale (VAS): mean value of 19 vs. 37]
11. Kraigher et al. (2005) [24]	110	PUS/3 weeks	Reduction in heroin (VAS: mean value of 46 at baseline vs. 15 at day 20) and cocaine craving (25 at baseline vs. 12 at day 20)
1. Kastelic et al. (2008) [14]	67	PUS/4 weeks	Decrease in craving for other drugs in both subgroups, especially for heroin and alcohol
10. Kraigher et al. (2002) [23]	67	PUS/3 weeks	Reduction in heroin (VAS: mean value of 48 at baseline vs. 15 at day 20) and cocaine craving (24 at baseline vs. 10 at day 20)
12. Vasilev et al. (2006) [25]	20	PUS/6 months	Reduction in the craving for heroin (VAS: mean value reduced from 7.7 ± 2.0 at baseline to 2.3 ± 2.3 at week 24)
Additional drug consumption
6. Eder et al. (2005) [19]	64	RCT/7 weeks	Number of injection sites decreased, cocaine usage increased over time and benzodiazepine consumption remained stable in both groups
9. Mitchell et al. (2003) [22]	14	CT/6 weeks	Similar for both treatments
11. Kraigher et al. (2005) [24]	110	PUS/3 weeks	Decrease in cocaine additional consumption (40% of patients had cocaine consumption at baseline compared with 23.3% at day 20)
1. Kastelic et al. (2008) [14]	67	PUS/4 weeks	Unchanged
10. Kraigher et al. (2002) [23]	67	PUS/3 weeks	Reduction in benzodiazepine consumption (52% at day 1 vs. 35% at day 20) but cocaine consumption remained unchanged
12. Vasilev et al. (2006) [25]	20	PUS/6 months	Reduction in the illicit use of heroin and methadone
13. Rao et al. (2005) [26]	74	PUS/4 weeks	Reduction in the heroin consumption (61% of patients reported that they had stopped heroin, 24.2% reported using heroin only for 5 days over the previous month)
4. Giacomuzzi et al. (2006) [17]	240	CSS	Lower consumption of cocaine and opioid in all three maintenance groups than in patients at admission
Adverse events
6. Eder et al. (2005) [19]	64	RCT/7 weeks	SROM 82% (toothache, headache, constipation and influenza) and methadone 76% (toothache, vomiting, headache, stomachache)
8. Mitchell et al. (2004) [21]	18	CT/6 weeks	SROM was associated with fewer and less troublesome adverse events than methadone (15 ± 25 vs. 62 ± 33)
12. Vasilev et al. (2006) [25]	20	PUS/6 months	Constipation (n = 5), sweating (n = 1)
1. Kastelic et al. (2008) [14]	67	PUS/4 weeks	Improvement in number of adverse events in subgroup A with significant adverse events related to methadone (4.2 ± 2.7 at day 0 and 2.3 ± 3.0 at day 28)
13. Rao et al. (2005) [26]	74	PUS/4 weeks	Gastritis (n = 3), pruritis (n = 2), constipation (n = 2), sedation (n = 1)

Abbreviations are as follows: CSS, cross-sectional study; CT, controlled trial; PUS, prospective uncontrolled study; RCT, randomized controlled trial; VAS, visual analogue scale.

^†Exact 95% confidence interval assessed by using binomial distribution.

In almost all uncontrolled studies, quality of life, withdrawal symptoms, craving and additional drug consumption were improved with SROM maintenance treatment. In controlled studies, similar improvements were obtained with methadone [15, 19].

Considering quality of life, well-being was improved with SROM [14, 25]. In addition, SROM was associated with higher SF-36 Social Functioning scores than methadone (mean SF-36 Social Functioning score of 78 ± 18 vs. 58 ± 25) [21]. In the comparative study, however, no statistically significant difference was found in any quality of life domain of the Berlin Quality of Life Profile between SROM and methadone [15].

Withdrawal symptoms were particularly reduced during the first 20 days of SROM maintenance [14, 23–25]. However, a switch between maintenance drugs in comparative studies led to a short recurrence of withdrawal symptoms [19, 21].

Craving for heroin was reduced significantly with SROM [14, 21, 23–25]. In the randomized controlled trial (methadone vs. SROM), craving for heroin decreased and remained low in both groups [19].

Some studies reported that SROM was associated with fewer and less serious adverse events, particularly in patients with significant adverse events related to methadone, but this applied mainly to open-label crossover studies [21].

Tables 4 and 5 detail the characteristics and the results of each of the included studies (study size, participants, intervention, comparison, outcomes, study design, follow-up period and main results, including P-values).

Table 4.

Characteristics of included studies (ordered by article ID)

Article	Study size	Participants	Intervention	Comparison	Outcomes	Study design	Follow-up period	Notes
1. Kastelic et al. (2008) [14]	67	Geographic region: Slovenia Opioid-dependent subjects undergoing methadone maintenance treatment with significant side effects related to methadone (A) or inadequate symptom control despite daily methadone dose ≥90 mg (B). Mean age: 29.5 years 73.1% male	7 days preswitch phase, 7 days to switch from methadone to SROM, 21 days of SROM maintenance phase [583 (A) and 989 mg daily (B)]	No control group Subgroup analyses: A vs. B	Retention rate, withdrawal symptoms, quality of life, craving, additional drug consumption, adverse events, global rating of efficacy and treatment preference	Prospective uncontrolled study	35 days	–
2. Winklbaur et al. (2008) [15]	64	Geographic region: Austria Patients with an opioid dependence without opioid maintenance therapy Mean age: 28.7 years 87.5% male	7 weeks of SROM followed by 7 weeks of methadone maintenance (A) or 7 weeks of methadone followed by 7 weeks of SROM maintenance (B)	SROM (680 mg daily) vs. methadone (85 mg daily)	Quality of life	Randomized controlled trial. Crossover design	14 weeks	Same sample as 6. Eder et al. (2005) [19]
3. Mitchell et al. (2006) [16]	14	Geographic region: South Australia Opioid-dependent volunteers receiving methadone maintenance treatment Mean age: 35 years 78.6% male	Switch from methadone (78 ± 32 mg) to SROM (349 ± 137 mg) during ∼6 weeks and then switch back to methadone	No control group	Pain sensitivity and mood status	Controlled trial. Crossover design	6–10 weeks	Same sample as 9. Mitchell et al. (2003) [22]Remuneration upon completion of the study
4. Giacomuzzi et al. (2006) [17]	240	Geographic region: Austria Opioid-dependent volunteers with or without opioid maintenance treatment Mean age: 26 years 60% male	–	Methadone (39 mg)/buprenorphine (9.3 mg)/SROM (235 mg)/no maintenance treatment (NMT)	Quality of life, withdrawal symptoms, additional drug consumption	Cross-sectional study	–	–
5. Giacomuzzi et al. (2005a) [18]	27	Geographic region: Austria Opioid-dependent volunteers receiving SROM or buprenorphine treatment Mean age: 30.2 years Percentage of men unknown	–	SROM (348.6 mg daily) vs. buprenorphine (8 mg daily)	Driving capacity	Cross-sectional study	–	Same sample as 7. Giacomuzzi et al. (2005b) [20] with different comparison groups
6. Eder et al. (2005) [19]	64	Geographic region: Austria Patients with an opioid dependence without opioid maintenance therapy Mean age: 28.7 years 87.5% male	7 weeks of SROM followed by 7 weeks of methadone maintenance (A) or 7 weeks of methadone followed by 7 weeks of SROM maintenance (B)	SROM (680 mg daily) vs. methadone (85 mg daily)	Retention rate, withdrawal symptoms, craving, additional drug consumption, depression, anxiety, physical complaints, adverse events	Randomized controlled trial. Crossover design	14 weeks	Same sample as 2. Winklbaur et al. (2008) [15]
7. Giacomuzzi et al. (2005b) [20]	41	Geographic region: Austria Opioid-dependent volunteers receiving SROM or methadone treatment Mean age: 29.7 years Percentage of men unknown	–	Methadone (59.1 mg daily) vs. SROM (348.6 mg daily)	Driving capacity	Cross-sectional study	–	Same sample as 5. Giacomuzzi et al. (2005a) [18] with different comparison groups
8. Mitchell et al. (2004) [21]	18	Geographic region: South Australia Opioid-dependent volunteers receiving methadone maintenance treatment Mean age: 36 years 50% male	Switch from methadone (78 ± 32 mg) to SROM (347 ± 131 mg)	No control group	Quality of life, withdrawal symptoms, patient rating of efficacy and acceptability, treatment preference, additional drug consumption	Controlled trial. Crossover design	6 weeks	Remuneration upon completion of the study
9. Mitchell et al. (2003) [22]	14	Geographic region: South Australia Opioid-dependent volunteers receiving methadone maintenance treatment Mean age: 35 years 78.6% male	Switch from methadone (78 ± 32 mg) to SROM (349 ± 137 mg)	No control group	Pharmacodynamics and pharmacokinetics criteria, withdrawal symptoms, additional drug consumption, treatment preference	Controlled trial. Crossover design	6–10 weeks	Same sample as 3. Mitchell et al. (2006) [16]. Remuneration upon completion of the study
10. Kraigher et al. (2002) [23]	67	Geographical region: Austria Patients with an opioid dependence without opioid maintenance therapy Mean age: 28.6 years 61.2% male	3 weeks of SROM maintenance (mean daily dose 593.4 mg)	No control group	Retention rate, additional drug consumption, craving, withdrawal symptoms	Prospective uncontrolled study	3 weeks	–
11. Kraigher et al. (2005) [24]	110	Geographical region: Austria Patients with an opioid dependence without opioid maintenance therapy Mean age: 27.9 years 80% male	3 weeks of SROM maintenance (mean daily dose 665 mg)	No control group	Retention rate, withdrawal symptoms, craving, additional drug consumption, somatic complaints	Prospective uncontrolled study	3 weeks	–
12. Vasilev et al. (2006) [25]	20	Geographical region: Bulgaria Patients with an opioid dependence without opioid maintenance therapy Mean age: 28.8 years 80% male	6 months of SROM maintenance (mean daily dose 760 mg)	No control group	Retention rate, withdrawal symptoms, craving, additional drug consumption, patients’ well-being, adverse events	Prospective uncontrolled study	6 months	-
13. Rao et al. (2005) [26]	74	Geographical region: India Patients with an opioid dependence Mean age: unknown Percentage of men unknown	At least 4 weeks of SROM maintenance (60–240 mg daily dose)	No control group	Retention rate, additional drug consumption, working activity, criminal activity (pick-pocketing, stealing), adverse events	Prospective uncontrolled study	4 weeks	–

Table 5.

Results of individual studies (ordered by study ID)

Article ID	Summary data for each group	Effects estimated
1. Kastelic et al. (2008) [14]	SROM maintenance	Retention rate: 80.6% (54 of 67) at 28 days.
	Subgroup analyses: • Significant side effects related to methadone (A): n = 39, mean age 29.4 ± 5.9 years, 71.8% male	Withdrawal symptoms (SOWS): A, 4.1 ± 3.7 at day 0 and 1.6 ± 3.0 at day 28 (P < 0.001); B, 9.0 ± 6.7 at day 0 and 2.9 ± 3.5 at day 28 (P < 0.001).
		Craving: decrease in craving for other drugs in both subgroups, especially for heroin and alcohol
	• Inadequate symptom control despite daily methadone dose of ≥90 mg (B): n = 28, mean age 29.8 ± 6.7 years, 75% male	Quality of life: SCL27, enhanced well-being in both subgroups. Global WHO QOL-BREF score detail: A, 52.7 ± 17.3 at day −7 and 64.8 ± 12 at day 28 (P < 0.001); B, 44.9 ± 25.2 at day −7 and 54.5 ± 17.5 at day 28 (P > 0.05).
		Additional drug consumption: unchanged in both subgroups.
		Adverse events: improvement in number of adverse events in subgroup A: 4.2 ± 2.7 at day 0 and 2.3 ± 3.0 at day 28 (P < 0.001). No data available for subgroup B.
		Global rating of efficacy: A, very satisfied 61%, satisfied 21%; B, very satisfied 39%, satisfied 39%.
		Global treatment preference: 96% (52 of 54) of patients preferred SROM maintenance treatment.
2. Winklbaur et al. (2008) [15]	SROM/methadone (A): n = 32, mean age 29.5 ± 7.5 years, 84.4% male, duration of heroin consumption 64.2 ± 52.7 months	Quality of life (BQLP): comparing group A and group B: no statistically significant difference in any quality of life domain (P from 0.10 to 0.97). Time effect: improvement in nearly all quality of life domains between baseline and week 14 [general well-being (P < 0.001), mental health (P = 0.001), general state of health (P = 0.018), leisure time at home (P = 0.034) and leisure time out of home (P = 0.008)].
	Methadone/SROM (B): n = 32, mean age 27.9 ± 5.6 years, 90.6% male, duration of heroin consumption 60.2 ± 49.2 months
3. Mitchell et al. (2006) [16]	Crossover: switch from methadone to SROM (during ∼6 weeks) and then switch back to methadone.	Pain sensitivity (cold pressor test, electrical stimulation test): pain responses were nearly identical for each drug.
	n = 14, mean age 35 years, 78.6% male	Mood status (Profile and Mood States, Morphine Benzedrine Group scale): overall mood disturbance did not differ between drugs.
4. Giacomuzzi et al. (2006) [17]	Methadone: n = 40, mean age 27.3 ± 6.4 years, 57% male	Quality of life (BQLP) – overall satisfaction, methadone 5.3 ± 1.5, buprenorphine 4.9 ± 1.4, SROM 4.1 ± 1.7, NMT 3.5 ± 1.6 (P < 0.001).
	Buprenorphine: n = 40, mean age 26.3 ± 7.5 years, 57% male	Withdrawal symptoms (Opiate Withdrawal Scale): both the buprenorphine and the methadone maintenance group showed less stomach cramp, muscular tension, general pain, feeling of coldness, heart pounding, runny eyes and aggression than patients at admission (P < 0.05).
	SROM: n = 40, mean age 27.8 ± 4.8 years, 57% male	Additional drug consumption: significantly lower consumption of cocaine and opioid in all three maintenance groups than in patients at admission (P < 0.001).
	No maintenance treatment (NMT): n = 120, mean age 25.3 ± 7.1 years, 63% male
5. Giacomuzzi et al. (2005a) [18]	SROM: n = 14, mean age 29.7 ± 6.5 years Buprenorphine: n = 13, mean age 31.1 ± 7.8 years	Driving capacity: better psychomotor performance in patients with buprenorphine maintenance compared with SROM within Visual Pursuit Test [correct answers (P = 0.016), working time (P = 0.047)].
6. Eder et al. (2005) [19]	SROM/methadone (A): n = 32, mean age 29.5 ± 7.5 years, 84.4% male, duration of heroin consumption 64.2 ± 52.7 months	Retention rate: A (SROM) 84.4% (27 of 32) at 7 weeks; B (methadone) 90.6% (29 of 32) at 7 weeks/odds ratio 1.79–95% confidence interval (0.39–8.22).
	Methadone/SROM (B): n = 32, mean age 27.9 ± 5.6 years, 90.6% male, duration of heroin consumption 60.2 ± 49.2 months	Additional drug consumption: the number of injection sites decreased, cocaine usage increased significantly over time in both groups (P < 0.001) and benzodiazepine consumption remained stable.
		Craving: craving for heroin, cocaine and alcohol decreased and remained low in both groups.
		Withdrawal symptoms (Wang scale): significant decrease after week 1, withdrawal recurring during the crossover phase.
		Depression, anxiety and physical complaints: patients receiving slow-release morphine had significantly lower depression (P < 0.001), anxiety scores (P = 0.008) and fewer physical complaints (P < 0.001).
		Adverse events (percentage of patients reporting at least one side effect): SROM 82% (toothache, headache, constipation and influenza), methadone 76% (toothache, vomiting, headache and stomachache).
7. Giacomuzzi et al. (2005b) [20]	Methadone: n = 27, mean age 29.7 ± 6.9 years SROM: n = 14, mean age 29.7 ± 6.5 years	Driving capacity: better psychomotor performance in patients treated with methadone compared with SROM within: • Vienna Reaction Test System: reaction time (P = 0.004), correct reactions (P < 0.001); • Tachistoscopic Traffic Test Manheim for Screen: more correct answers (P < 0.001); • Cognitrone test: lower working time (P < 0.001), lower time for correct reactions (P < 0.001).
8. Mitchell et al. (2004) [21]	Crossover: switch from methadone to SROM (during ∼6 weeks) and then switch back to methadone.	Quality of life: SROM was associated with higher SF-36 Social Functioning scores than methadone (78 ± 18 vs. 58 ± 25, P = 0.006).
	n = 18, mean age 36 years, 50% male	Withdrawal symptoms: withdrawal severity during the first 5 days of SROM maintenance was greater than during the resumption of methadone maintenance (P = 0.02).
		Treatment preference: 77.7% of subjects preferred SROM to methadone (14 of 18).
		Treatment efficacy and acceptability: compared with methadone, SROM was associated with fewer and less troublesome adverse events (15 ± 25 vs. 62 ± 33, P = 0.001), greater drug liking (P = 0.01) and reduced heroin.
		craving (VAS: mean value of 19 vs. 37, P = 0.03).
		Additional drug consumption: no other significant differences between methadone and SROM.
9. Mitchell et al. (2003) [22]	Crossover: switch from methadone to SROM (during ∼6 weeks) and then switch back to methadone.	Pharmacodynamics and pharmacocinetics criteria: opioid effects were of a similar overall magnitude following dosing for each drug and showed an inverse association with plasma drug concentrations, which peaked later for morphine compared with methadone (P < 0.001).
	n = 14, mean age 35 years, 78.6% male	Opioid withdrawal severity: no significant differences between drugs (P = 0.58).
		Additional drug consumption: similar for both treatments (P > 0.05).
		Treatment preference: 85.7% of subjects preferred SROM to methadone (12/14, P = 0.01).
10. Kraigher et al. (2002) [23]	SROM: n = 67, mean age 28.6 years, 61.2% male	Retention rate: 94% (63 of 67) at 3 weeks.
		Withdrawal symptoms: decreased with SROM maintenance (mean Wang scale value of 11.2 at day 1 compared with 1.1 at day 20).
		Additional drug consumption: reduction in benzodiazepine consumption (52% at day 1 vs. 35% at day 20, P < 0.01); cocaine consumption remained unchanged.
		Craving: reduction in heroin (VAS: mean value of 48 at baseline vs. 15 at day 20, P < 0.0001) and cocaine craving (24 at baseline vs. 10 at day 20, P < 0.0001).
11. Kraigher et al. (2005) [24]	SROM: n = 110, mean age 27.9 years, 80% male	Retention rate: 93.6% (103 of 110) at 3 weeks.
		Withdrawal symptoms: decreased at day 20 compared with baseline (mean Wang scale value of 12.1 at day 1 compared with 0.3 at day 20, P < 0.0001).
		Additional drug consumption: decrease in cocaine additional consumption (40% of patients at baseline had cocaine consumption compared with 23.3% at day 20, P = 0.0083).
		Craving: reduction in heroin (VAS: mean value of 46 at baseline vs. 15 at day 20, P < 0.0001) and cocaine craving (25 at baseline vs. 12 at day 20, P < 0.0001).
		Somatic complain: the incidence of somatic complaints was reduced (Clinical Self Rating Score of the Munich Psychiatric Information System: mean value of 21.7 at baseline vs. 12.5 at day 20, P < 0.0001).
12. Vasilev et al. (2006) [25]	SROM: n = 20, mean age 28.8 years, 80% male	Retention rate: 95% (19 of 20) at 6 months.
		Withdrawal symptoms: SOWS scores reduced markedly in the first 2 weeks and remained low until week 24 (mean value of 16.9 ± 6.6 at baseline vs. 6.6 ± 5.9 at week 24).
		Additional drug consumption: reduction in the illicit use of heroin and methadone.
		Craving: reduction in the craving for heroin (VAS: mean value reduced from 7.7 ± 2.0 at baseline to 2.3 ± 2.3 at week 24, P < 0.001).
		Patient well-being: improvement of patient well-being from baseline assessment; improvements with regard to suicidal depression (85%), anxiety and dysphoria (66%), general illness (58%), social dysfunction (54%), sense of hopelessness (34%), attention (25%), self-reported typical depressive (27%) and disease-related symptoms (11%).
		Adverse events: Five episodes of constipation and one episode of sweating.
13. Rao et al. (2005) [26]	No control group	Retention rate: 44.6% (33 of 74) at 4 weeks.
	After 4 weeks of SROM maintenance: n = 33, mean age 39 years, 100% male	Additional drug consumption: 61% of patients reported that they have stopped heroin, 24.2% reported using heroin only for 5 days over the previous month.
		Working activity: proportion of patients working regularly increased from 21.2 to 57.6%.
		Criminal activity: decreased from 45 to 3%.
		Adverse events: gastritis (n = 3), pruritis (n = 2), constipation (n = 2) and sedation (n = 1).

Values are presented as means ± SD. SROM, slow-release oral morphine; SOWS, Short Opiate Withdrawal Scale; SCL27, Symptom Check List 27; WHO QOL-BREF, WHO Quality of Life-BREF; BQLP, Berlin Quality of Life Profile; VAS, visual analogue scale.

Discussion

Data about SROM efficacy are scarce in the literature. Few studies considering the use of SROM for OMT are available, and most of them were uncontrolled or not randomized. Moreover, it is not possible to conclude that SROM is equivalent to methadone in enhancing quality of life, reducing craving, controlling withdrawal symptoms or decreasing additional drug consumption.

Our systematic review suffers from several mandatory limitations. First, a reporting bias is possible. In order to minimize this risk, an active search was performed to find all studies dealing with SROM for OMT, multiple databases were searched and references cited in selected articles checked. One study providing preliminary data about SROM for OMT in patients intolerant to methadone was not included in qualitative analyses because only those patients completing the maintenance programme were assessed (16 patients, retention rate of 100%) and because no standardized measuring instruments were used to assess psychopathological features [27]. However, this study showed that adverse events related to methadone could be ameliorated after the switch to SROM maintenance (particularly weight gain). Second, a selective reporting of complete studies or outcomes within studies is also possible (publication bias). In order to limit the risk of missing data about outcomes within studies, a comparison of outcomes listed in the methods section of the published reports was made with those for which results were presented. Finally, risk of publication bias of nonsignificant findings or uncompleted studies was limited by requesting trials registries (Clinical Trial.gov, Cochrane Central Register of Controlled Trials). No supplementary study was retrieved by this method.

Compared with methadone or buprenorphine, SROM maintenance efficacy is not well documented, and the available studies are too heterogeneous to be analysed together. A meta-analysis published in 2008 comparing methadone and buprenorphine maintenance included 24 randomized and controlled clinical trials [28]. We found only one randomized controlled study with a crossover design comparing SROM with methadone [19].

Because of the lack of control groups in most of the studies, it is difficult to know which part of the measured effect is because of the drug or of the psychosocial and psychotherapeutic interventions. Psychosocial and psychotherapeutic interventions combined with pharmacotherapy have been shown to be effective in treatment outcome studies [29]. These approaches aim to increase treatment motivation, prevent relapse and reduce harm. In addition, they may provide advice and practical support to patients who have to address their housing, employment and family-related problems in parallel with treating their opioid dependence.

Retention in treatment was the main assessment criterion in previous published meta-analyses comparing efficacy of different drugs for OMT [28, 30]. In our review, retention rate was available in only six of the nine included. It ranged from 95% at 6 months to 80.6% at 4 weeks. In addition, the follow-up period was short in most of these studies (between 3 and 7 weeks). Randomized and controlled trials assessing methadone and buprenorphine efficacy were longer. For example, trials performed by Johnson et al. [31] and Mattick et al. [32] found, respectively, a 17-week retention rate of 73% for high-dose methadone and 58% for buprenorphine, and a 13-week retention rate of 59% for methadone and 48% for buprenorphine.

The demographic characteristics of patients included in these studies (mean age ranging from 26 [17] to 39 years [26]) were similar to those in the population of opiate users under maintenance treatment in Occidental countries. In Europe, the mean age of patients entering outpatient treatment for primary opioid use is 33 years, and almost all countries have reported an increase since 2003. An average time lag of about 8 years was reported between first use of opioids and first drug treatment [6].

While defining pharmacological criteria for the use of drugs in the treatment of opioid dependence, Montastruc et al. qualified SROM as a poorly satisfying drug for OMT [33]. In fact, SROM exhibited only two of the six defined criteria (same pharmacodynamic properties as the drug being substituted and compatibility with a socially satisfying quality of life), missing out on four of them (duration of action ≥24 h, few euphoric and reinforcing effects, oral or sublingual administration without any special affinity for intravenous routes, clinical randomized comparative trials and security data allowing a New Drug Application for the drug as an OMT). This review has confirmed these results, especially concerning the lack of randomized comparative trials.

However, clinical points of view about SROM maintenance differ greatly among physicians. Although successful experience in the treatment of heroin-dependent patients with long-acting morphine was reported [34], other data discuss the real efficacy of SROM as an OMT. Risk of misuse is often discussed; some reports suggest that most SROM prescription is diverted to i.v. injecting [35], with the risk of microembolisms because of insoluble contents, such as talcum [36]. Moreover, a recent study suggested that the number of deaths related to morphine may have increased after the introduction of SROM as an OMT in the Austrian regions of Tyrol and Vorarlberg [36]. In Europe, expansion in maintenance drugs for opioid-dependent users has been accompanied by increasing reports of misuse of these drugs. This is not only true for SROM but also for buprenorphine and methadone, the most diverted drugs in 2009 [6].

Slow-release oral morphine is one of the most powerful drugs available for management of severe and chronic pain. Extending indications of SROM to OMT will increase the risk of misuse and could compromise legitimate use in pain management. This could partly explain the lack of pivotal clinical trials to assess SROM efficacy for OMT. However, a recent multicentre, multinational phase III study with a crossover design is underway, comparing the effectiveness of SROM vs. methadone maintenance treatment in patients who have previously been treated with methadone [37]. With more than 300 patients followed up over a 6 month period, this study will provide highly relevant data about SROM efficacy for OMT.

Slow-release oral morphine was initially studied for OMT of pregnant addicted women, with the hypothesis that SROM may produce less severe neonatal abstinence syndrome than methadone [38, 39]. A study conducted by Fischer et al. [39] did not find any reduction in the number of days that neonatal abstinence syndrome was experienced by neonates with SROM compared with methadone and concluded that both drugs were suitable maintenance agents for pregnant opioid addicts.

A recent randomized controlled trial investigated SROM efficacy for opioid detoxification [40]. No statistically significant differences were found between SROM and methadone in terms of completion rate (proportion of patients completing the study), signs and symptoms of opioid withdrawal, craving for opioid or self-reported symptoms.

Slow-release oral morphine is now available in several European countries as an alternative to methadone (Austria, Slovakia, Slovenia, Bulgaria and Luxembourg) [8]. The preparation commonly used is Substitol® (morphine sulfate pentahydrate; Mundipharma AG, Basel, Switzerland). In Austria, the Narcotics Maintenance Decree revised in 1998 defined methadone as the first choice. Slow-release oral morphine can be used by specialist clinics for opioid maintenance in the event of strong adverse drugs reactions or in the case of pregnancy or HIV infection [35]. In France, methadone and buprenorphine are two drugs fully approved for maintenance therapy. Slow-release oral morphine can be prescribed exceptionally in cases of methadone or buprenorphine failure [9]. More scientific data about SROM efficacy are needed in order to revise legislation about SROM in France.

Conclusion

In conclusion, as most of the studies assessing SROM efficacy were uncontrolled, there is now no definite evidence showing that SROM is an effective alternative to methadone for OMT. Further research should be planned in order to investigate the clinical utility of this drug for this kind of indication.

Competing Interests

There are no competing interests to declare.