1 June 2007

Randomised Swedish study shows buprenorphine suits 35% versus methadone at 80%

American Journal of Psychiatry 2007 164;5:797-803

A Stepped Care Strategy Using Buprenorphine and Methadone Versus Conventional Methadone Maintenance in Heroin Dependence: A Randomized Controlled Trial. Kakko J, Gr�nbladh L, Svanborg KD, von Wachenfeldt J, R�ck C, Rawlings B, Nilsson L-H, Heilig M.

This prominent Swedish group may have done the field another service without realising it. A previous buprenorphine trial by Kakko et al. (Lancet 2003) had a placebo group with so many deaths (25% within a year) that they unintentionally demonstrated the life saving benefits of buprenorphine treatment (for the first time to my knowledge). They also showed that, at least in the Swedish environment with restricted access to alternative maintenance treatments, even second line treatment can yield a high retention rate of 75% at one year.

In the present study, half of 96 addicted subjects received standard methadone treatment (mean dose 111mg daily) while the rest received buprenorphine with transfer to methadone if needed. The latter occurred according to defined criteria of drug use or cravings, but only after a buprenorphine dose escalation to 32mg daily if this was tolerated. Intensive psychosocial support was also provided. Treatment was supervised daily in the first month when the trial was still double-blinded, in what these authors themselves describe as the �sensitive initial month of treatment�. After documented stability, there was graduation to twice and even once weekly supervised medication with the remainder as �take-away� or dispensed doses, consistent with good practice (see Rhoades 1998).

The final mean dose of buprenorphine was 29mg daily (�5). Consistent with previous research, only 35% of buprenorphine patients remained on the treatment (42% transferred to methadone while 21% dropped out, most notably in the first 3 weeks).

The authors claim that retention rates in the two groups were �virtually identical� (77% bup, 79% meth), yet their decay graph clearly shows an excess of buprenorphine drop outs in the first three weeks of treatment, which is consistent with the reported experience of others. There were only one or two early drop outs in the methadone group yet in those randomised to start with buprenorphine, six or more subjects appear to have departed the trial in this period. Methadone had about the same number of people leaving treatment overall but they did so at a more constant rate over the 6 months of the trial. The high retention rates are probably due to both the high quality methadone treatment offered and lack of alternative sources of maintenance treatment in Sweden.

Both groups had approximately 80% clear urine tests for opiates, also consistent with the adequacy of doses and high level of psychosocial treatment provided these subjects.

Despite these unsurprising findings, the authors write in the article�s abstract: �strategies using buprenorphine as a first line treatment should be considered�. This is changed, without additional explanation or references, to a much stronger conclusion in the text that buprenorphine �should generally be used as the first-line treatment in heroin dependence, with provisions for rapid progression to methadone when needed�. It seems bizarre to recommend a drug with a known failure rate of over 60%, especially when there was an acceptable, cheaper and more effective alternative. In addition, methadone is considered the drug of choice in pregnancy.

We are told that the combination buprenorphine/naloxone drug was chosen partly because it has been approved in the United States. From a �pure science� point of view, this clouds the issue since most research is on the pure product with very little on the combination product. The authors speculate that the addition of naloxone might have caused the high average dose of buprenorphine at 29mg daily (�5mg). The reference cited supporting dose equivalence (Johnson RE 2000) reports no such comparison from my reading of the original paper. The finding of high doses required is consistent with Bell et al 2004 (which is NOT cited by these authors, despite being the only clinical comparison to date, so far as I know). In this, patients needed approximately 50% more buprenorphine when transferred from pure buprenorphine to the combination product. They detail the low primary toxicity of buprenorphine yet they ignore some major problems reported with buprenorphine treatment (deaths in France, diversion in Victoria, injection, low efficacy, costs). They also fail to remind readers that most methadone overdose has occurred from its use outside the supervised clinic setting, meaning pain management prescriptions, break-ins, etc. See Ballesteros where only 4% of methadone overdose deaths in North Carolina in a five year period occurred in registered maintenance treatment recipients.

We are informed that this trial was partly funded by Schering-Plough Sweden. Four of the authors have received funds from drug companies. Three of these were from the manufacturer of buprenorphine and one author received funds from this source in three separate countries (Estonia, Sweden and Australia).

These authors quote the term �non-inferiority� four times in their article regarding their customised �stepped care strategy� based on retention rates and toxicology. In this comparison, however, they do not take into account several other important factors including dose, costs and patient satisfaction. They state that �no patients commented on transitioning� (sic) from buprenorphine/naloxone to methadone. Yet such patients must have been faring poorly on buprenorphine and must have had some views on finally being prescribed the more effective drug. Some might also possibly have had some views at being given the less effective drug initially in the trial, despite the consent process. Have any of these �transitioning� patients been asked to comment on the authors� suggestion relating to first line buprenorphine, I wonder? One may worry that some needle sharing or overdose might have occurred in the interim period, not to mention the 12.5% who appear to have dropped out altogether in this early period from the �stepped care� group. We are not informed if any died although in Sweden such information would probably be easily available, as in previous reports.

It would also be instructive (and equitable) to know how many of the methadone patients may have clinically indicated a transfer to buprenorphine. We find the rate is between 10 and 30% in our own practice. This may well have applied to some of the drop-outs and even to some of those retained in the treatment, who may have preferred buprenorphine for a variety of reasons. Such a matched or parallel protocol might have increased methadone retention to 90% or higher. Hence this is another flaw in the current trial and another reason to question the authors� enthusiastic endorsement of the buprenorphine combination as a first line drug in dependency treatment.

Note that an editorial by Kathleen Brady in the same issue points out several more disadvantages of the current study. Few medical practices would have the resources to give such intensive psychosocial support as occurred here. Brady uses the incorrect term �naltrexone� rather than �naloxone� no less than ten times. She also states that if injected, the �naltrexone will precipitate withdrawal�. Assuming that she is referring to naloxone, this is not the case for most buprenorphine maintained patients and only for a proportion of others with current habits on pure opiate agonists (eg. methadone, heroin or morphine). She also points out that the �stepped care� can only work if there is closely supervised treatment such as in existing methadone clinics. This would defeat the supposed purpose of office based treatment which has not been extensively tested against clinic based treatment to my knowledge. I believe that the authors are also mistaken in their choice of the term �stepped care�. This would refer to additional medication or other treatment, based on pre-existing criteria. This trial, however, uses a �safety net� or �rescue� protocol and not a �stepped care� approach in my view.

Comments by Andrew Byrne ..


Ballesteros MF, Budnitz DS, Sanford CP, Gilchrist J, Agyekum GA, Butts J. Increase in Deaths Due to Methadone in North Carolina. JAMA 2003 290:40

Bell J, Byron G, Gibson A, Morris A. A pilot study of buprenorphine-naloxone combination tablet (Suboxone�) in treatment of opioid dependence. Drug Alcohol Rev (2004) 23;3:311-318

Kakko J, Svanborg KD, Kreek MJ, Heilig M. 1-year retention and social function after buprenorphine-associated relapse prevention treatment for heroin dependence in Sweden: a randomised, placebo-controlled trial. (2003) Lancet 361:662-668

Nielsen S, Dietze P, Dunlop A, Muhleisen P, Lee N, Taylor D. Buprenorphine supply by community pharmacists in Victoria, Australia: perceptions, experiences and key issues identified. Drug Alc Review 2007 26;2:143-152

Rhoades HM, Creson D, Elk R, Schmitz J, Grabowski J. Retention, HIV Risk, and Illicit Drug Use during Treatment: Methadone Dose and Visit Frequency. 1998 Am J Public Health 88:34-39