Commentary on Reimer et al. (2011): More work needed


In 1898 Bayer marketed their newly synthesized opioid diamorphine in the United States as a cure for morphine addiction. Without the benefit of randomized controlled trials (RCTs), observers came to believe that heroin had limited effectiveness in curing addiction, and the drug fell into disfavour. Are we at last getting the evidence that Bayer needed? Is heroin-assisted treatment (HAT) a good treatment for addiction—or, at least, better than methadone maintenance therapy (MMT)? This appears to be the conclusion of the authors of this study [1], who write: ‘These data support the superiority of heroin compared to methadone in severely opioid-dependent subjects’.

What is ‘severe opioid dependence’? This study recruited subjects unable to take advantage of the respite from compulsive drug use offered by MMT, and either persisted in heroin use while in treatment or dropped out of treatment. Such individuals are often demoralized and marginalized. A sense of hopelessness, sometimes described as a lack of ‘motivation to change’, makes patients ‘hard to treat’. One of the few incentives motivating them to engage in treatment is access to prescribed heroin. The current report illustrates that in addition to suppressing street heroin use, remaining in HAT was associated with improved health—and that self-reported and staff-rated improvements were generally slightly greater in the heroin than the methadone group.

Prospective subjects wanted diamorphine, and 29% of subjects randomized to methadone declined treatment, apparently preferring to remain on their low-structure, low-dose MMT—taking enough to block withdrawal, but not so much as to abolish the reinforcing effects of heroin. The availability of such ineffective treatment is counter-therapeutic, and has contributed to giving MMT a poor reputation. HAT should be reserved for those who fail structured, adequate dose MMT. HAT is not a competitor to MMT, but part of a stepped-care treatment system, in which the possibility of eventual access to HAT may encourage people to accept MMT.

Because diamorphine is a more rewarding and motivating treatment, and treatment was not double-blind, self-reported improved quality of life by participants must be regarded circumspectly. There is also a distinct possibility of staff bias influencing outcomes. In the multi-site German trial, while all clinics achieved close to 70% retention in diamorphine subjects, retention in the methadone group varied widely, from 20 to 75% [2]. Treatment works better when clinicians believe in it, and one plausible explanation for the discrepancy is that, in some centres, staff were enthusiastic about diamorphine, and felt that those randomized to methadone had missed out on good treatment.

The price of better retention in HAT was a significantly higher incidence of adverse events, respiratory depression or seizures immediately mainly post-injection. The authors conclude that all patients should be kept for 15 minutes post-injection; but just as there is no single version of MMT, so HAT has evolved differently in different settings. Where the German trial used very low ancillary doses of methadone (8 mg/day was the mean), our UK clinic uses considerably higher methadone doses, and observes administration of methadone as well as diamorphine. In addition, the clinic has a ‘no benzodiazepine’ policy, and nurses are vigilant in monitoring for intoxication pre-injection. Since instituting these policies, during more than 60 patient-years of treatment we have not had a hypoxic episode post-injection. HAT clinics need trained and vigilant staff, monitoring of patients after dose changes but, depending on their dosing policies, do not need to keep all patients for 15 minutes post-injection.

Can a patient addicted to heroin give a meaningful assessment of whether their life is better when they are prescribed the drug? I think they can, mainly, but a critical appraisal of heroin prescription also needs objective data, including objective data on the medical consequences of HAT. The current report cites improvement in one objective measure, body mass index, in both groups, with no difference between treatments (challenging the folklore that ‘methadone makes you fat’). A Swiss pilot study identified a number of potentially important side effects of diamorphine treatment, particularly changes in sex hormones [3]. The medium and long-term health effects of HAT, and the optimal model of treatment, remain to be studied adequately.

Declaration of interests

Dr Bell is in receipt of research funding from ReckittBenckiser, which markets a range of buprenorphine products