Commentary on Burns et al. (2015): Retention in Buprenorphine Treatment

The paper in this issue by Burns et al.(1) uses a large administrative dataset to show that among individuals with opioid use disorder in Australia those treated with buprenorphine had poorer retention in treatment than did those treated with methadone. These important and somewhat concerning findings add to a growing body of evidence that, indeed, rates of treatment retention with buprenorphine fall below rates observed with methadone.

A recent Cochrane Review performed a meta-analysis of 5 double-blind, randomized, controlled trials (N=788) that compared flexible dose buprenorphine to flexible dose methadone and found significantly inferior retention with buprenorphine.(2) This review further determined that in trials with fixed low doses of the two medications, patients on buprenorphine also had poorer retention but observed no differences in retention with fixed moderate or high doses. Nevertheless, the authors concluded that, since flexible dosing better reflects clinical practice, methadone promotes better retention than buprenorphine.

Furthermore, a large, multisite, randomized, open-label comparison of buprenorphine to methadone published subsequent to the Cochrane review which used flexible dosing but had high average doses showed significantly worse retention with buprenorphine.(3) However, higher buprenorphine doses were associated with better retention.

An evaluation of electronic medical record data from 34 treatment clinics in the U.S. also observed better retention with methadone than with buprenorphine.(4)

These studies occurred in opioid treatment programs where patients had to attend frequently and received observed dosing. What about buprenorphine retention in office-based settings? Reports from office-based settings also portray suboptimal retention rates ranging from 66% at 14 weeks to 59.6% at 12 months.(5-7)

How important is treatment retention? It probably represents the most important outcome in the treatment of opioid use disorder. We know that this disorder has a chronic relapsing course. Relapse typically ensues rapidly upon discontinuation of medication-assisted treatment,(7-9) and mortality rates also jump at this juncture.(10, 11) Criminal behavior and infectious disease risk increase as well when patients drop out.(12, 13)

What steps might be taken to ameliorate apparent inadequate rates of retention in buprenorphine treatment? Buprenorphine dosage offers one initial point of leverage. The paper of Burns et al.(1) notes lack of dose information as a limitation, but data from the Cochrane Review(2) and the multisite trial mentioned above(3) suggest that higher doses may improve retention. Many physicians prescribing buprenorphine believe that most patients stabilize on 12mg to 16mg per day,(14) despite findings from human laboratory research that full blockade of heroin effects usually requires higher doses.(15) The field needs a blinded, randomized dose ranging study that would compare outcomes, particularly treatment retention, on buprenorphine doses of 16mg, 24mg, 32mg, and possibly 48mg daily. Careful measurement of buprenorphine and nor-buprenorphine plasma levels during the study would help to determine medication adherence and underscore any dose response effects seen.

Within the pharmacologic realm pharmacogenetics, though it has thus far proven elusive, may offer opportunities for improvement in retention. For example, DNA collection occurred for a subsample of participants enrolled in the multisite trial mentioned above.(3) Single nucleotide polymorphisms in the gene coding for the delta-opioid receptor were associated with differential treatment response (albeit illicit opioid use, not treatment retention) among African-Americans(16) and European American women.(17) If these provocative findings were replicated, and additional pharmacogenetic predictors discovered, such biomarkers might help match patients to the optimal medication to keep them in treatment.

Moving beyond pure pharmacology we need to understand the overall buprenorphine treatment experience from the patient's perspective. Then we can design patient-centered and supportive programming to interdigitate seamlessly with the pharmacologic intervention, perceive quickly when the patient hovers on the verge of leaving treatment, and intervene to prevent dropout. This effort will require largely qualitative research. An initial attempt at doing so was conducted with participants who dropped out of the study cited above.(3, 18) Some participants indicated that they did not get adequately informed about buprenorphine before receiving it. Some had a negative induction experience with ongoing withdrawal symptoms even days after the first dose. Some had persistent side effects. Some were actually dissatisfied that the medication made them feel “normal” and did not provide a euphoric effect. Transportation issues and lack of time to attend treatment figured into dropout for some. Based upon these data, as first steps we would want to provide more in-depth patient education about the medication, be cautious and attentive during the induction period, remain alert to and address side effects, and mitigate practical barriers to receiving buprenorphine.

Spurred on by the work of Burns et al. and others, we as a field have an urgent imperative to improve retention on medication assisted treatment, particularly buprenorphine. We need more research to help determine optimal dosing strategies, to figure out who will most likely respond well to buprenorphine, and to ascertain how to make the entire treatment experience more conducive to extended patient engagement.