Addiction (2005) 100: 1090-1100
A multi-center randomized trial of buprenorphine-naloxone versus clonidine for opioid detoxification: findings from the National Institute on Drug Abuse Clinical Trials Network. Ling W, Amass L, Shoptaw S, et al.
Dear Readers,
This 14 day trial showed buprenorphine-naloxone prescribed patients significantly more likely to stay in treatment and stop using street drugs than those taking clonidine. It also found much better completion rates for in-patients (77%; 22%; n = 77; 36) compared with out-patients (29%; 5%; n = 157; 74) prescribed the opioid versus clonidine. Considering the buprenorphine was given in therapeutic doses of 2mg daily up to the last day, these results are not surprising.
However, it is intriguing that there are so many inconsistencies in this report from such highly respected authors. Most useful randomised trials compare an intervention of known utility with a protocol of promising but uncertain benefit - using parallel protocols and clearly defined, positive end points regarding the subjects' health. There are usually equal numbers of participants but it appears this trial was halted early by a pre-determined protocol due to the poor outcomes of the clonidine group. Unusual for 'Addiction', this multi-centre trial does not appear to have a statement concerning the funding (although the buprenorphine combination tablets were supplied by the manufacturer). Nor could I see a statement on conflict of interest. In-patients and out-patients are included but selection criteria are not given, nor are they randomised. Also unusual for Addiction, there is more than half a page of acknowledgements containing 85 names and up to 20 institutions. Several aspects of the article remind me of the sort of item contained in sponsored journal 'supplements'. One wonders if the normal peer-review process occurred, so numerous are the inconsistencies, omissions and questions concerning a lack of specific clinical aims/focus in the piece.
As above, most of the findings are predictable from the literature. The study involved a comparison of two non-evidence based 'treatments', being clonidine for detoxification and buprenorphine combination for 'detoxification'. Both 'treatments' have been shown to be ineffective in trials with over 90% failure rates on conservative criteria at medium term follow-up (1 to 3 months). Thus this study, whatever its findings, is unlikely to add to the scientific literature or knowledge base. We know that prescribing buprenorphine in standard doses to heroin addicts attracts and retains subjects while also reducing heroin use. We also know that clonidine does not do any of the above to any measurable or consistent extent although its use still has supporters and there is some promise of its chemical cousin, lofexidine (Brit-Lofex) for withdrawal symptoms. As yet, we still do not know whether the buprenorphine combination (bup-combo) drug is as effective as pure buprenorphine. There are indications that it is less efficacious (Bell) although another paper (Strain) showed contrary-wise that the combination drug caused consistently higher blood levels of 10-40% and thus might be expected to yield a better response. Thus readers may ask why use a combination drug rather than the pure drug, especially when the combination drug is not approved for commencement of treatment. The combination drug's only claimed benefit is reduced attractiveness to the street market. Yet dose diversion would not seem to be an issue here, being largely supervised and only short-term.
Unsurprisingly, the authors find that a higher proportion of bup-combo opioid dependent patients remained in 14 days of prescribed "detox" than those given no opioids. Also, and most importantly, they found that in-patients fared much better than out-patients in all respects (apart from hapless pair who perished in the in-patient group).
The authors fail to clearly define 'opioid detoxification'. Some might find the prescribing of opioids during 'detoxification' incongruous. Opioid detoxification cannot really commence until the patient ceases all significant intake of opioids. Even on day 13, these patients were offered 2mg of buprenorphine, still ten times the standard analgesic dose of 0.2mg sublingual (the maximum dose given - on day 3 - was 16mg). A further period of 14 days follow-up for the bup group might have allowed a fairer comparison of the two approaches in this case.
Either way, the authors have not clearly defined what they mean by detoxification, nor have they explained their statement in the abstract that 'success' involved an 'opioid-free urine sample on the last day of clinic attendance'. Perhaps their pathologist was unable to detect buprenorphine. Later in the article they state this must be free of *illicit* opioids, without defining *illicit*. Buprenorphine, like methadone, morphine or codeine can be licit or illicit. Indeed, it was to address the issue of diversion that the combination drug was introduced into the USA. Yet we now know that it can be abused, and that there is little if any evidence of less diversion than for other prescribed opioids.
In several countries now, even heroin can be legally prescribed and it is possible that, prescribed in decreasing doses, it might have comparable or better results than buprenorphine reductions. The very next item in Addiction is, in fact, a favourable comparison of oral morphine with methadone from experienced researchers in Austria.
Adverse events: The section on side effects starts out with a long, confusing, 'de Quincey-esque' sentence of 62 words. In the following 57 lines comes a tedious and confusing 'lecture' to the reader on the technical difference between 'serious adverse events' and 'adverse events'. Only at line 37, and after the reassuring statement 'Few serious adverse events occurred in either protocol' do we first learn of two deaths in this 14 day, 344-subject study. This implies a very high average mortality and requires some detailed explanation. All we are told is that one death (in the bup-combo group) was due to 'respiratory failure' (could this be 'code' for drug overdose?) and one was from the clonidine treated group ('bacterial endocarditis' - usually a slow or sub-acute disease in drug addicts). We are told (without coronial references) that 'neither death was attributed to study medication'. However, we are also told that patients in both groups took an average of three other medications during the trial, including 'OTC' acetaminophen (paracetamol), ibuprofen, loperamide and diphenhydramine as well as prescribed oxazepam, lorazepam, phenobarbital, hydroxyzine, methocarbamol, trimethobenzamide, Donnatal (containing phenobarb), zolpidem, trazadone and doxepin.
I have never heard of a patient dying whilst undergoing supervised detoxification from opioids, so these mortality reports need to be taken very seriously, especially when the patients were in a NIDA approved treatment protocol in registered and (presumably) accredited medical facilities. I can only speculate (as the authors do not) that this was a trial which attracted the most severely ill American addicts and/or may not have provided the best means to rehabilitation for their condition. Neither trial protocol is an 'evidence based' or proven modality for heroin addiction, even though both may be reasonable choices for subjects as long as they also have the option of more effective avenues such as traditional methadone maintenance. In many American (and Australian) cities there are still long waiting lists and/or methadone is beyond the financial reach of many addicts.
The authors state "In the out-patient group, 18 serious events occurred, with 14 in the bup-nx group and four in the clonidine group." Equal numbers of serious events occurred in the in-patient groups, including two with suicidal behaviour and one with persistent vomiting in those receiving bup-nx. I am mystified how an in-patient can have a 'motor accident' whilst under treatment, yet, along with 'spine surgery' we are offered this as an 'adverse event' without further explanation. It is also hard to understand how a patient who was assessed as suitable for detoxification could have died so quickly of sub-acute bacterial endocarditis. Detoxification is highly unwise in those with active septicaemia. One patient dropped out due to 'sensitivity to a study medication' (unspecified). This 'unfortunate' omission becomes a 'serious' one if buprenorphine were implicated.
The most important finding of this trial is probably that in-patient detoxification was so much more efficacious than community treatment (despite there being no randomisation). The differences were dramatic and they point up the inappropriate decisions made by successive administrations to close down drug and alcohol detoxification wards in New South Wales and elsewhere. Much of this was based on claims by so-called experts of their services not being 'cost-effective' or 'evidence based'. This trial shows that by using counselling and the Californian "Matrix" protocols and a 'placebo' such as clonidine, excellent in-patient detox results can be obtained in selected patients.
Comments by Andrew Byrne ..
Conflict of interest: Dr Byrne earns a proportion of his income dispensing methadone and buprenorphine to registered dependency patients.
Please note that Professor Charles O'Brien has written an excellent editorial in this 'Addiction' pointing out many of the above problems. Were he or I a reviewer, this month's edition might have been thinner.
