JAMA 2005; 294:903-913
Collins ED, Kleber HD, Whittington RA, Heitler NE. Anesthesia-Assisted vs Buprenorphine- or Clonidine-Assisted Heroin Detoxification and Naltrexone Induction - A Randomized Trial.
Dear Colleagues, The current [August 2005] JAMA home page features this item with the caption: "Anesthesia-Assisted Heroin Detoxification: Collins and colleagues randomly assigned treatment-seeking heroin-dependent patients to anaesthesia-assisted rapid detoxification with naltrexone induction, buprenorphine-assisted rapid detoxification with naltrexone induction, or clonidine-assisted opioid detoxification with delayed naltrexone induction. They found that withdrawal severity, treatment completion and retention, and proportions of opioid-positive urine specimens during 12 weeks of outpatient treatment were comparable across the 3 methods of detoxification."
In fact, there are five items in this edition mentioning naltrexone, two being letters to the editor on a long acting injectable form for alcoholism. The current feature has a commentary by Patrick O'Connor on the role of detoxification. There is also is a glowing historical tribute to Vincent P. Dole and colleagues whose seminal methadone report was published in this same summer holiday edition of JAMA 40 years ago.
With veteran researcher Herbert Kleber, this group from Columbia describe a randomised comparison of naltrexone induction in heroin addicts using three methods: (1) rapid detox under 4-6 hour anaesthetic (2) buprenorphine bolus and (3) clonidine with traditional in-patient detoxification.
The study raises several important ethical questions while also giving perhaps the last word on rapid detoxification from opioids, a century after the first report [MacLeod, N. Cure of morphine, chloral, and cocaine habits by sodium bromide. BMJ (1899) 15/4/1899 p896]. The main results are unremarkable: viz (a) that almost 100% of anaesthetised patients successfully take their first dose of naltrexone, (b) that rapid detox is hazardous, (c), that the naltrexone "treatment" is of very limited benefit (75-90% of subjects dropped out by 12 weeks) and (d) that the particular method of detoxification has no significant impact on rates of medium-term abstinence.
It is possible that some side effects in the anaesthesia group (n=35) may reflect this team's lack of experience as well as their limited ability to elicit a clear history from their patients. All three subjects who developed major anaesthetic complications are said to have had "concealed" histories (of diabetic ketoacidosis, bipolar disorder, pneumonia and sleep apnoea) from the researchers. To ascribe each anaesthetic complication to deceitful subjects is rather unusual and there may be alternative views. Others have reported lower complication rates, yet there is no doubt that such treatment can be hazardous in this population, especially if they come directly from street heroin habits.
Prescription of naltrexone for opioid addiction as a 'treatment' has only little limited scientific support in unselected candidates in community treatment. Some believe that it may have benefits in carefully selected subjects (as stated by O'Brien in the same issue p888). So why did these authors go to so much trouble to 'induct' addicts into an ineffective treatment? I note that some providers now give very frank details about the expected success rates of their treatments. Yet others have claimed '100% success' rates and call their detoxification treatments 'painless'.
It is predictable that those taking pure buprenorphine were retained for slightly longer than those given clonidine, which may be little more than a placebo in this situation. And it is self-evident that the anaesthesia patients were more likely to take their first dose of naltrexone which is given while they are still unconscious.
In his accompanying editorial Patrick O'Connor tells us that over 3 million Americans have used heroin and ten times that number prescribed opioids. Even more worrying is that over 1% of school children in the US had used heroin in 2004. Thus we are dealing with an epidemic in anyone's terms.
As with McGregor and Ali's randomised study from Adelaide (D&A Rev) rapid detox shows no significant benefits over traditional detoxification in the medium term (3, 6 or 12 months). In view of the high risks and poor results, there should probably be no further studies of rapid detoxification in unselected subjects. It is still possible that longer acting forms of naltrexone may yet prove effective for those seeking abstinence. Formal research on the safety and effectiveness of such novel delivery methods is awaited.