Treatment of prescription opioid dependence: main results and longterm follow-up outcomes from the Prescription opioid addiction treatment study (POATS) Roger D. Weiss, M.D. Professor of Psychiatry,Harvard Medical School Chief, Division of Alcohol and Drug Abuse, McLean Hospital, Belmont, MA March 11, 2016

Background Prescription opioid (PO) epidemic

 2nd most commonly abused drug  More overdose deaths than heroin & cocaine combined  Most studies of opioid dependence had focused on heroin users in MMT  Can’t assume same findings with PO users

Background • Buprenorphine was the most likely treatment option to consider for Rx opioid users • Proven efficacy for treatment of opioid dependence • More flexible and convenient than methadone due to office-based prescription • Some questions remain about optimal delivery (e.g. role of counseling)

Main POATS design Wei

Overview of POATS

 Largest randomized clinical trial yet conducted for prescription opioid dependence

 Conducted at 10 sites across the U.S.

 N=653  2-phase adaptive treatment research design

Research questions

 Does adding individual drug counseling to buprenorphine-naloxone tx + standard medical management improve outcome?  What length of bup-nx tx is best for prescription opioid dependence?  1 month vs. 3 months?  Maintenance?  Does current chronic pain or a history of heroin use affect outcome?

Phase 1, up to 12 weeks

Bup detox with or without drug counseling

Phase 2, 24 weeks

Bup maintenance with or without drug counseling

Main study results Wei

Successful outcome 60

SMM only

50

SMM + Drug counseling

40 % 30 20 10 0 Phase 1: Taper

Phase 2: Maintain

Week 16: End of taper

Week 24: Follow-up

Phase 1 vs. Phase 2, maintain: p<.001; Phase 1 vs. Phase 2, follow-up: p=.21; Phase 2, Maintain vs. Follow-up: p<.001

The role of heroin use & chronic pain

Did lifetime heroin use at baseline (23%) predict successful outcome? % of pts with successful outcomes

60 50

Heroin use

40

None

30 20 10 0

*p<.003

End of bup tx Week 12*

8 weeks post-taper Week 24

Phase 2

Did chronic pain at baseline (42%) predict successful outcome? % of pts with successful outcomes

60 50

Chronic pain

40

No chronic pain

30 20 10 0 End of bup tx Week 12

8 weeks post-taper Week 24

Phase 2

Primary reason for use in chronic pain patients, past & present  1° reason for 1st use  Pain  Get high  1° reason for current use IF 1st reason was pain  Avoid withdrawal  Pain  Get high Weiss et al. J Subst AbuseTreatment 2014

83% 13%

56% 23% 14%

Long-term follow-up (LTFU) study: Rationale & design

LTFU study rationale

 No previous longitudinal studies of treatment-seeking PO dependent patients

 Nothing is known about  the long-term course &  response to treatment of this population

LTFU study design

 Aim: to examine long-term outcomes of participants in the main trial  Naturalistic, exploratory study  Participants were interviewed at 18, 30, & 42 months post-initial randomization  Telephone interviews: 45-60 minutes each  $75 for each interview completed  All interviews conducted by McLean Hospital staff

LTFU Sample Description N=375 enrolled N =338 interviewed

Sample description (N=375) Demographics

 44% female  Mean age = 33  90% white  48% never married  64% employed full-time Psychiatric history, past-year

 Major depressive disorder:  PTSD: 12%

21%

Sample description continued Substance use history at baseline

 16% past-year non-opioid SUD  22% ever used heroin  32% previous opioid use disorder treatment  63% 1st used opioids to relieve pain  48% used opioids ≥4 years

Illicit opioid use, dependence and abstinence

Mean days using drugs, past 30 days, over time 30

Baseline Month 18 Month 30 Month 42

25 20 15 10 5

0

p<.001

Rx opioids*

Marijuana*

Opioid dependence over time, pastmonth (excluding agonist therapy) 50

100% at baseline 40

30

% 20

16 12

10

8

0

Month 18

Month 30

Month 42

p<.02

Illicit opioid abstinence past-month by current opioid agonist treatment 100

%

80

Agonist tx 80

Not

60

40

80

51

37

20

0

*p<.001

Month 18*

Month 42*

Treatment utilization over time

Opioid agonist treatment over time, past-month 50

40

38

37

Month 30

Month 42

32 30

% 20

10

0

Month 18

SUD treatment, past year Treatment, % (n) None Buprenorphine maintenance Self-help groups Outpatient counseling Medical detoxification Methadone maintenance Inpatient/residential Intensive outpt/day hospital Naltrexone

Months 18, 30, & 42 34-39% 41-43% 35-41% 23-32% 6-12% 6-9% 3-7% 2-3% 0-1%

Initiation of risk behaviors

Prevalence of new heroin users >4 days, past month

Follow-up

New heroin users

Month 18 (n=252)

11 (4%)

Month 30 (n=312)

16 (5%)

Month 42 (n=306)

0

Prevalence of injection heroin use >5 times, past year

Follow-up

1st injection use

Month 18 (n=252)

12 (5%)

Month 30 (n=312)

18 (6%)

Month 42 (n=306)

4 (1%)

Deaths

 2 participants died during main trial: 1 during Phase 1, 1 during Phase 2

 5 participants (1.5%) died during LTFU period

Predictors of Long-term Outcomes • Those successful in treatment in the main trial were more likely to be abstinent at 18month follow-up, but this was no longer the case by Month 42 • Those with lifetime heroin use before entering the study were more likely to be opioid dependent at follow-up

Conclusions Wei

Conclusions Encouraging results

 Rates of opioid abstinence in LTFU surpassed main trial extended bup-nx treatment outcomes  Especially if receiving agonist treatment (80%)  Even if not on agonist treatment, rates at month 42 matched outcomes on bup-nx at 12 weeks (50% successful)  Decreased opioid use was not accompanied by increased use of other substances.

Conclusions Cautionary results  New heroin & injection use  8% of participants reported 1st use of heroin during LTFU  10% reported injecting heroin for the 1st time, >5x in past year  1.5% died during LTFU  History of any heroin use prior to study entry associated with poor outcomes

Limitations

 Generalizability: 52% response rate

 LTFU participants may be more or less severe

 Interviews

 During tx trial, in person + urine drug screens  In LTFU, by telephone without urine drug screens

 Main trial success less strict measure than complete abstinence required for successful LTFU outcome

Conclusions

 1st longitudinal study examining treatmentseeking patients with prescription opioid dependence  Opioid agonist treatment effective in supporting long-term recovery, though some recover without agonist treatment  Initiation of heroin use &/or injection can occur  Should be monitored over the course of agonist treatment

Acknowledgements

McLean POATS LTFU team,

with

3 World Series trophies

Our home base: Proctor House, McLean Hospital

Acknowledgements, cont’d

• • • •

Investigators, clinicians, & patients at the 10 study sites Jennifer Potter, Ph.D., M.P.H. and Walter Ling, M.D. CCTN, CCC, and DSC-1 staff NIDA Grants U10DA15831 & K24DA022288