Download SIH

Supervised heroin treatment for
refractory chronic heroin addicts:
development, research study and
clinical provision in the UK
John Strang
National Addiction Centre (The Maudsley & Institute of Psychiatry)
London
(on behalf of RIOTT research, clinical and related colleagues)
Declaration - general

* DH, NTA, Home Office, NACD, WHO, UNODC

* Diamo, Reckitt-Benkiser, Schering-Plough, GenusBritannia, GW, Napp, Titan, Catalent, Auralis, Mundi

* Phoenix House, Clouds House, Lifeline, KCA, Action
on Addiction, Society for the Study of Addiction
RIOTT funding support & declarations

Research Funding
 Community Fund (Big Lottery) & Action on Addiction & Hedley Foundation

Clinical Services Funding
 National Treatment Agency, Department of Health, and Home Office
 Local DATs & PCTs

Medications:
 Diamo, Switzerland; Cardinal, UK; Auralis, UK; also Genus, UK

Other support
 The Band Trust – DVD
 EMCDDA – European analysis and ‘Insights’ report

Clinical colleagues:
 Marina House, Maudsley; Darlington; Brighton

Service users/patients/study subjects:
RIOTT Team & Collaborators

Investigators/trial coordination
 Prof John Strang
 Dr Nicholas Lintzeris
 Dr Nicola Metrebian

Local Investigators
 Dr Deborah Zador / Dr James Bell
 Dr Tom Carnwath/Dr Soraya Mayet
 Dr Hugh Williams


Research staff
 Vikki Charles
 Luciana Forzisi
 Teodora Groshkova
 Chris Hallam
 Anthea Martin

RIOTT clinical team leaders
 Rob van der Waal, London
 Anne McNutt, Darlington
 Ian Wilson, Brighton

Trial co-ordination
 National Addiction Centre, Institute of
Psychiatry, KCL

Statistician
 Laura Potts, Clinical Trials Unit,
Institute of Psychiatry, KCL

Health Economics
 Dr Sarah Byford Institute of
Psychiatry, KCL
 Barbara Barrett, Institute of
Psychiatry
Clinical Trial Pharmacist
 Glynis Ivin, Maudsley Hospital
 Godwin Achunine, London clinic
Diamorphine suppliers
 DiaMo Narcotics GmbH, Switzerland
 Auralis, UK
Randomisation
 Clinical Trials Unit, IoP

Pathology
 Dr Andy Marsh & Richard Evers,
Kings College Hospital
Structure of today’s talk

history of heroin policy; and new scrutiny
 The
RIOTT trial – origins, conduct and results
 The
trial - and dialogue with government
 European
 Ongoing
advances
current analyses
Sir Humphry Davy Rolleston,
(President of Royal College of Physicians, 1922-1936)
Sir Humphry Davy Rolleston,
(President of Royal College of Physicians, 1922-1936)
 The
legitimacy and authority of the medical
versus law enforcement perspective
 “maintenance”
(not termed thus) with injectable
morphine or diamorphine (heroin) legitimate
medical practice
 Sets
UK apart from post-1920s US policy
CHANGES IN THE UK IN THE 1970s

initial optimism for therapeutic power in the
new drug clinics post-1968; but then growing
disillusionment over the next decade or so
 The
growing status of oral methadone
 The
withering of injectable heroin
 Intermediate
years of injectable methadone
WHAT INJECTABLE PRODUCTS?
(n.b. predominantly an English phenomenon)
 Two
-
products:
heroin ampoules
(dry amps) (less than 1%)
-
methadone ampoules
(wet amps) (approx 10%, now
maybe 1%)
(historically also morphine by injection)
Methadone: ampoules as proportion
proportion of NHS methadone prescriptions as
Ampoules (England, annual data, 1990-2008)
10
% as Amps
9
8
7
6
5
4
3
2
1
0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
Structure of today’s talk

history of heroin policy; and new scrutiny
 The
RIOTT trial – origins, conduct and results
 The
trial - and dialogue with government
 European
 Ongoing
advances
current analyses
To complement the development of existing services,
heroin should be available on prescription
to all those who have a clinical need for it.
The number of people receiving heroin will increase as
overall numbers in treatment grow.
The administration of prescribed heroin for
those with a clinical need will take place in
safe, medically supervised areas with clean
needles. Strict and verifiable measures will
be in place to ensure there is no risk of
seepage into the wider community.
UK Government Drug Strategy, 2002
Unsupervised vs Supervised

‘Old’ (unsupervised)

Long history

But minimal research
evidence base

Internationally isolated

Mainly for the stable

‘New’ (supervised)
Supervised vs unsupervised

‘Old’ (unsupervised)

‘New’ (supervised)

Long history

Increasingly strong
research evidence base

But minimal research
evidence base

In line internationally

Internationally isolated

Public safety

Mainly for the stable

Accords with Drug
Strategy 2002 & 2008

For the repeatedly
‘failing’
My starting observations
 The
‘Old British System’ of injectable maintenance
and the new supervised treatment are extremely
different.
 The
evidence base for ‘Old British System’ is
extremely weak scientifically (although not
necessarily negative).
 The
evidence base for ‘Swiss-style’ supervised
injectable maintenance (as used in all recent
RCTs) is increasingly strong.
Accumulating body of evidence

((Hartnoll et al, 1980, Archives Gen Psych – UK))

Perneger et al, 1998, BMJ – Switzerland

Van den Brink et al, 2003, BMJ – Netherlands

March et al, 2006, JSAT – Spain

Haasen et al, 2007, B J Psych - Germany

Oviedo-Joekes et al (NAOMI), 2009, NEJM - Canada

Strang et al (RIOTT), 2010, Lancet; and ongoing - England
Operating costs

……..

Optimised oral methadone maintenance – c 5k pppa

Supervised injectable methadone maintenance – c 10k pppa

Supervised injectable heroin maintenance – c 15k pppa

…..
Operating costs

‘bog-standard’ oral methadone maintenance – c 3k pppa

DTTO/DIP methadone treatment + monitoring – c 10k pppa

Optimised oral methadone maintenance – c 5k pppa

Supervised injectable methadone maintenance – c 10k pppa

Supervised injectable heroin maintenance – c 15k pppa

Prison – c 44k pppa
Operating costs
 ‘An
ineffective service is inefficient and
cannot be cost-effective, no matter how
cheaply it is provided’
• Cochrane, 1972
Supervised injecting clinics
Characteristics of new clinics
7
days per week; under supervision
 no
take-home injections / adequate daily doses
 oral
take-home supplements
 flexible
prescribing - oral take-home conversion on
request
 dedicated
facility - specific function
What was the aim & design
of the trial?
Target population
Entrenched heroin addicts who have
repeatedly been found to fail to
benefit from existing treatments
(despite treatment, continuing to inject
heroin on all/most days per month)
Second-line use of injectable maintenance
Rx-seeking
dependent
heroin user
Treat with oral
good-quality
maintenance
repeated
treatment ‘failure’
Poor benefit with
oral maintenance
‘Optimisation box’
still treatment
‘failure’
minimal
benefit
Still poor benefit
with oral
Brief test trial of
‘RIOTT’
treatment
Good benefit
Immersion in full
‘RIOTT’
treatment
Computer generated randomisation
Injecting heroin
User in opioid
Maintenance
Treatment for
6 months
Diamorphine iv/im
+/- oral methadone
Methadone
Ampoules iv/im
+/- oral methadone
Enhanced
Oral
Methadone
What were our measures of
effective treatment?
Primary outcome measure
Primary outcome
Measures
Reduction in street heroin The proportion of subjects in each
use
group who cease regular street
heroin use
Metabolism of “illicit” Heroin
Diamorphine
Noscapine
Papaverine
HO
CH3
HO
O
CH3
O
N
H3C
O
O
H3C
N
O
O
O
CH3
N
O
O
O
CH3
CH3
O
H3C
HO
Codeine
H3C
O
Meconine
6-Monoacyl morphine
6-Hydroxypapaverine
HO
HO
CH3
O
N
H3C
O
O
O
HO
N
CH3
O
OH
HO
O
H3C
Morphine
6- Desmethylmeconine
4,6-Dihydroxypapaverine
Outcome measures
Secondary outcomes
Measures
Other illicit drug use
UDS & self-report
Treatment retention
Clinic records (& self report)
Injecting practices
Frequency, risk & complications
Psychosocial functioning & Quality
of Life Measures
SF-36, EQ-5D, OTI
Crime
Self-report (drug related
expenditure & criminal activity)
Safety
Adverse events
Patient satisfaction
Semi-structured Q’s
Cost effectiveness
Service costs (internal & external)
How many patients were
retained in treatment?
0
10 20 30 40 50 60 70 80 90
100
Retention
0
2
4
6
8
10
12
14
16
18
20
22
24
26
Weeks
Injectable
Methadone
Injectable
Heroin
Oral
Methadone
28
Treatments to be investigated
Supervised Injectable Heroin (SIH)
Supervised Injectable Methadone (SIM)
Optimised Oral Methadone (OOM)
Sample to be analysed
Intention-To-Treat (ITT) sample
Per-Protocol (PP) sample
Primary outcome
Retention in treatment Χ
Reducing/quitting ‘street heroin’
Other drug use; well-being;
Criminal behaviour ?
Wider recovery
‘responder’ or ‘abstinent’?
Major reduction in frequency of use
of ‘street heroin’
Completely abstinent from ‘street
heroin’
Which measure of primary outcome?
Urine test results
Observations and measurements
Self-report
Types of urinalysis datasets
‘raw’ data (actual clinical results)
Data incl. imputations (enables analysis)
Data incl. imputations adjusted for
stratification (ensures no inadvertent bias)
What doses are prescribed and
how quickly is the new treatment
established?
daily diamorphine dose (mg)
RIOTT- doses for the three groups (OOM, SIM, SIH)
daily diamorphine dose (mg)
RIOTT- doses for the three groups (OOM, SIM, SIH)
daily diamorphine dose (mg)
RIOTT- doses for the three groups (OOM, SIM, SIH)
daily diamorphine dose (mg)
RIOTT- doses for the three groups (OOM, SIM, SIH)
What were the benefits?
To begin at the end
Four important conclusions, as I see them
•
SIH (heroin) group strongest achievement
•
SIM (inj methadone) better than control group
•
OOM (optimised oral) – notable benefit
•
Rapid onset of benefit and gain
So what are the main findings on
(i) ‘responder’ (reduced use of street-heroin)?
(ii) ‘abstinent from street-heroin’?
RIOTT - data on ‘responders’ and ‘non-responders’ –
broken down as % - at baseline (OOM, SIM, SIH)
100%
non-responder
90%
80%
responder
70%
60%
50%
100
100
100
0
0
0
OOM
SIM
SIH
40%
30%
20%
10%
0%
RIOTT treatment group
RIOTT - data on ‘responders’ and ‘non-responders’ –
broken down as % - at Months 4-6 (OOM, SIM, SIH)
100%
90%
27
80%
70%
60%
responder
72
67
50%
40%
73
30%
20%
10%
non-resp - some clean
28
33
OOM
SIM
0%
SIH
RIOTT treatment group
RIOTT - data on ‘responders’ and ‘non-responders’ –
broken down as % - at Months 4-6 (OOM, SIM, SIH)
non-responder
100%
90%
27
responder - only one dirty
responder - all clean
80%
70%
60%
72
67
50%
54
40%
30%
20%
10%
0%
31
19
7
2
OOM
SIM
SIH
RIOTT treatment group
RIOTT - data on ‘responders’ and ‘non-responders’ –
broken down as % - at Months 4-6 (OOM, SIM, SIH)
non-responder
100%
90%
27
80%
70%
60%
72
responder - > one dirty
responder - only one dirty
responder - all clean
67
35
50%
40%
19
30%
20%
10%
0%
19
24
7
7
2
OOM
SIM
19
SIH
RIOTT treatment group
RIOTT - data on ‘responders’ and ‘non-responders’ –
broken down as % - at Months 4-6 (OOM, SIM, SIH)
non-responder
100%
90%
27
80%
70%
60%
72
67
50%
16
responder - >2 dirty
responder - only 2 dirty
responder - only one dirty
responder - all clean
19
40%
19
30%
20%
17
24
10%
2
7
0
7
2
19
OOM
SIM
SIH
0%
RIOTT treatment group
So how substantial a benefit
are we talking about?
Odds ratios for >50% ‘heroinabstinent’ urines at 6/12 (ITT)
Factors
Category
Odds
ratio
Treatment
SIH vs OOM
8.2
(2.9, 23.2)
<0.001
SIM vs OOM
1.8
(0.7, 4.8)
0.25
SIH vs SIM
4.6
(1.7, 12.2)
0.002
95% CI
p-value
Odds ratios for completely ‘heroinabstinent’ urines at 6/12 (ITT)
Factors
Category
Odds
ratio
Treatment
SIH vs OOM
6.11
(1.93,20.44)
0.003
SIM vs OOM
2.31
(0.62, 8.54)
0.210
SIH vs SIM
2.65
(0.95, 7.38)
0.063
95% CI
p-value
The NNT calculation:
(Number-Needed-to-Treat)
NNT
SIH vs OOM
2.1
SIM vs OOM
9.1
SIH vs SIM
2.8
How quickly does this marked
advantage show itself?
Percentage of participants not using
illicit heroin by week (ITT sample)
Percentage of participants not using
illicit heroin by week (ITT sample)
Percentage of participants not using
illicit heroin by week (ITT sample)
Other outcomes
Retention in treatment
Other drug use
Well-being
Serious Adverse events
Criminal behaviour
Serious Adverse Events
OOM
9 SAE
SIM
4 SAE
9 unrelated
0 related
1 related
(1 x O/D)
1 in 5551
injections
SIH
7 SAE
2 related
(2 x O/D)
1 in 6613
injections
5 unrelated
3 unrelated
How real an issue? SAEs
 Injected
diamorphine –
 2 x rapid overdose requiring emergency naloxone as
well as oxygen (incl. unconscious and unrousable)
 Injected
methadone –
 1 x rapid overdose requiring emergency naloxone plus
oxygen
Oxygen saturation: IV versus IM
IM
96
IV
SpO2 (%)
94
92
90
0
10
20
30
40
Minutes post-injection
50
60
Oxygen saturation: IV versus IM
IM
96
IV
SpO2 (%)
94
92
90
0
10
20
30
40
Minutes post-injection
50
60
Oxygen saturation: case study
96
SpO2 (%)
93
90
87
Male, age 49
Intravenous diamorphine (6 years)
This dose = 120 mg
84
Daily dose = 400mg
0
10
20
30
40
Minutes post-injection
50
60
Structure of today’s talk

history of heroin policy; and new scrutiny
 The
RIOTT trial – origins, conduct and results
 The
trial - and dialogue with government
 European
 Ongoing
advances
current analyses
“… rolling out the prescription of injectable
heroin and methadone to clients who do
not respond to other forms of treatment,
subject to the findings, due in 2009, of
pilots exploring the use of this type of
treatment”.
(H.M.Government Drug Strategy, 2008)
RIOTT Research conclusions
Four important conclusions, as I see them
•
SIH (heroin) group strongest achievement
•
SIM (inj methadone) better than control group
•
OOM (optimised oral) – notable benefit
•
Rapid onset of benefit and gain
RIOTT Clinical conclusions
And four important clinical conclusions, also
•
Intensive-care – high-dose, high-level care
•
High-risk – be prepared
•
The most severe cases (?5-10%)
•
International critical mass with supervised
injectable maintenance treatment modality
Structure of today’s talk
 history
of heroin policy; and new scrutiny
 The
RIOTT trial – origins, conduct and results
 The
trial - and dialogue with government
 European
 Ongoing
advances
current analyses
HAT trials to date
Main paper
Hartnoll et al., 1980
Country
England
Perneger et al., 1998
Switzerland
Van den Brink et al.,
2003 (a and b)
March et al., 2006
Netherlands
Haasen et al., 2007
(a and b)
Germany
Oviedo-Joekes et al.,
2009 (+)
Strang et al., 2010
(+)
Spain
Canada
England
Key features of the new approach

Consistently (almost entirely, but not completely) now
considered as second-line treatment

Direct medical or nursing supervision of all injectable
doses
The supervised injecting heroinprescribing clinics

7 days per week (Spain week-days only); typically 23 blocks of opening hours

Higher daily doses; no take-home injections

Oral take-home supplements

Flexible prescribing - oral take-home conversion on
request

Dedicated facility - specific function
Mean treatment dose mg/day*
Country
SIH (+OM)
OM
Switzerland
509
n/a
Netherlands
548 (+71)
60
Spain
275 (+43)
105
Germany
442 (+8)
99
Canada
366 (+34)
96
399 (+46)
107
England
* trial data
Current features HAT practice
Country
Number
of clinics
Total
capacity
(range)
Catchment
area
Routine
practice
Switzerland
23
1,454
(15-210)
German-speaking part (22 clinics)
& Geneva (1 clinic)
Yes
Netherlands
17
650
(20-75)
Country-wide
Yes
Spain
1
56
(17 in Rx, July 2010)
Granada city
No
Germany
7
300
(12-70)
7 towns across Germany
Yes
Canada
2
140
Vancouver and Montreal
No
England
3
100
(30-40)
SE London, Brighton, and the NorthEast of England
Yes
Denmark
3
300
(25
in tx, July 2010)
Municipalities of Copenhagen,
Odense, and Glostrup
Yes
Retention
Perneger et
al., 1998
Van den
Brink et al.,
2003
March et
al., 2006
Haasen et
al., 2007
OviedoJoekes et
al., 2009
Strang et
al., 2010
Switzerland
Netherlands
Spain
Germany
Canada
England
Time to
follow up
6 months
12 months
9 months
12 months 12 months 6 months
SIH
(O)OM
93%,
92%
72%,
85%
74%,
68%
67%,
40%

88%,
54%
88%,
69%
Patients receiving HAT appear to be more likely to
be retained in treatment than those engaging in
oral methadone substitution therapy. This effect,
however, varied across the trials.
Outcome – abstinence or nearabstinence (a)
Study or Subgroup
van den Brink et al. 2003
Perneger et al., 1998
Strang et al., 2010
SIH (+OM)
OM
Events Total Events Total Weight
19
76
11
98 40.8%
21
27
7
21 33.4%
36
43
6
42 25.8%
Total (95% CI)
146
161 100.0%
Total events
76
24
Heterogeneity: Chi² = 4.52, df = 2 (P = 0.10); I² = 56%
Test for overall effect: Z = 5.78 (P < 0.00001)
Risk Ratio
M-H, Fixed, 95% CI
2.23 [1.13, 4.39]
2.33 [1.23, 4.41]
5.86 [2.76, 12.44]
Risk Ratio
M-H, Fixed, 95% CI
3.20 [2.16, 4.75]
0.1 0.2
0.5 1
2
5 10
Favours OM Favours SIH (+OM)
RIOTT Research conclusions
Four important conclusions, as I see them
•
SIH (heroin) group strongest achievement
•
SIM (inj methadone) better than control group
•
OOM (optimised oral) – notable benefit
•
Rapid onset of benefit and gain
RIOTT Clinical conclusions
And four important clinical conclusions, also
•
Intensive-care – high-dose, high-level care
•
High-risk – be prepared
•
The most severe cases (?5-10%)
•
International critical mass with supervised
injectable maintenance treatment modality
 Thank
you