May 2nd, 2016
Recently the neuroscientist Marc Lewis, a friend of mine, gave Percy Menzies, president of treatment provider Assisted Recovery Centers of America, the chance to guest-post on his blog as part of a series on the “response to the heroin epidemic.” Menzies’s post suggested that naltrexone, an opioid antagonist, is a “highly favorable alternative” to opioid addiction.
I cannot accept this message. The promotion of naltrexone as a viable solution for heroin users is not based on data, and when it is promoted as an alternative to opioid substitution therapy (OST), it is dangerous, unethical and a violation of human rights. Here are the facts.
Opioid Substitution Therapy Works
The first and most important thing to note is that opioid substitution therapy works. For methadone we have 50 years of irrefutable proof that it works. We use buprenorphine because it is shown to compare favorably to methadone in some robust clinical trials (1). In some cases methadone is better, in some cases buprenorphine—the pharmacological action differs—but they both work effectively (2) and any risks are easily outweighed by the benefits.
OST is shown to reduce mortality by up to 75 percent, reduce crime, improve health, improve retention in treatment and allow people the space to resolve many of the other issues that have made drug use so meaningful to them, perhaps at the expense of other, more beneficial activities. OST also reduces the spread of HIV(3) and is cost-effective.
I feel it’s essential to add that some people use OST because they like opioids. They enjoy them and want to keep using them, but that basic human right has been removed by our ridiculous prohibitionist policies. So to mitigate the harms of uncertain dose, quality and arrest, they choose to use methadone for most of the time and heroin at other times. I have no problem with this, and neither should anyone else.
Some people also have limited access to medical services, and while the privileged access scripts and prescription medications, they use street analogues. This may not be the majority, but they exist and need to be acknowledged.
So Why Isn’t OST Alone Solving the Problem?
Heroin levels remain high, despite the existence of OST, for three reasons: Firstly, OST is not widely available and secondly the manner in which it is offered is not welcoming to many heroin users. The third reason is that OST does not provide the same “high” as heroin and is not as attractive to those who would prefer to access diacetylmorphine, or pharmaceutical heroin. This should also be available, and the Swiss have demonstrated just how helpful it is.
Between 2003 and 2013 the use of OST as first-line treatment for people presenting with heroin use disorders declined from 33-27 percent in the US. While 45 percent of recipients were older than 45, only 22 percent were aged 20-34 (4). Just 3 percent of all US physicians are able to prescribe buprenorphine.
The lack of access to OST has been widely reported. For example, the Huffington Post ran a whole series called “Dying to be Free.” Disturbingly, US drug-court judges have even ordered people to stop their OST.
The Federal Guidelines for Opioid Treatment Programs (SAMHSA 2013) describe what are essentially abstinence-based, high-threshold programs with comprehensive admissions screening; they see OST as part of a “recovery orientated treatment plan,” and encourage “substance abuse” counseling, 12-step programs and mandatory continued drug testing through urinalysis (eight times per year minimum).
All of this discourages active heroin users from accessing OST services.
What Are Naltrexone and Extended-Release Naltrexone?
The website of Percy Menzies’ Assisted Recovery Centers of America (ARCA), describes their approach: Patients are detoxed “with the help of medications such as Suboxone® (buprenorphine) to relieve withdrawal symptoms. After this first step, you will be ready to begin taking the latest anti-craving medications, such as Vivitrol® (XR-NTX), or ReVia® (naltrexone) in our outpatient program and continue your daily life with few interruptions.”
What are these apparently miracle medications?
Naltrexone has been available since 1984 in pill form for treating opioid addiction; the extended-release injectable version has been available since 2010. Naltrexone is an opioid antagonist: In other words it has high affinity with the opioid receptor but unlike agonists (heroin or methadone) and partial agonists (buprenorphine), it has no effect.
Because of this high affinity, when an agonist like heroin is also ingested, the effect is blocked. The theory goes that if you are taking naltrexone, taking opioids is a pointless exercise. By blocking the normal functions of opioid use, the Pavlovian learning established through cue and reward dissipates over time.
In addition to this opioid-blocking function, ARCA describes the formulations of ReVia (in pill form) and Vivitrol (the 28-day injectable form) as anti-craving medications.
So, if promoters of naltrexone are to be believed, we have a medication that blocks the effects of opioids and relieves craving, and this should be offered far more widely in response to the US “heroin epidemic.”
But what does the data say?
The Data on Naltrexone
Various literature reviews show that naltrexone has little effect on levels of heroin use (5) compared to placebo. Compliance and retention are cited as the reason for this lack of effect (6). As would be expected in early stages of resolving a heroin use disorder, attitudes and motivations towards abstinence fluctuate. Unless closely monitored, people may decide to skip doses, creating the opportunity for heroin use. Retention in treatment is often a measure of efficacy in and of itself, and based on this alone, naltrexone does not compare favorably with OST (7). A Cochrane review shows no increased treatment retention for psychosocial interventions by adding naltrexone (8).
The longer acting version of naltrexone was developed in order to address these issues of compliance. XR-NTX (the 28-day-acting Vivitrol injection) guarantees compliance for a month at a time, with the individual is making an irreversible pre-commitment to abstinence. Through this pre-commitment to a longer-term antagonist, the individual mediates the effects of short-term cravings because they can’t be resolved through heroin use. However, this is often “tested” and it has been suggested that this “testing” of the blockade through heroin use is what improves abstinence (9).
As for cravings, studies have been unconvincing. Retention rates alone show us that daily-dose naltrexone does not address craving. OST does—hence its far higher retention rates. This is not surprising, as naltrexone is an antagonist. In some studies with XR-NTX, results have shown a perceived reduction in craving, although closer examination shows that craving diminishes with abstinence independently of naltrexone (10).
So naltrexone is no better than placebo unless compliance is enforced, and it has no effect on craving. And that’s the good news.
Naltrexone and Mortality
The biggest danger of naltrexone, both as oral daily dose or a 28-day injectable, is the substantial risk of overdose.
To be clear, during initiation methadone is also not without risks, and all OST, when terminated, presents a risk due to reduced tolerance. But naltrexone seems to increase that risk (11).
In one study, the risks with naltrexone were found to be between 2.8 and 7.4 times higher than with methadone (12). This study has been criticized by some, but what’s in little doubt is that naltrexone reduces tolerance. It is well known that antagonist use leads to upregulation of opioid receptors—meaning that receptors become super-sensitive—and therefore those that stop using naltrexone are at significantly increased risk of overdose.
Who Could Benefit?
All that said, we should not simply dismiss naltrexone—there is enough data to suggest that some recipients can do well. The UK National Institute for Health and Care Excellence (NICE) concludes its 2007 guidelines for naltrexone for the management of opioid dependence as follows:
“In summary, the Committee was convinced of the clinical effectiveness of naltrexone treatment in a selected, highly motivated group of people. The Committee concluded that for people who preferred an abstinence programme, who were fully informed of the potential adverse effects and benefits of treatment, and who were highly motivated to remain on treatment, naltrexone treatment would fall within acceptable cost-effectiveness limits.”
These sentiments are echoed by the World Health Organization. Their 2009 guidelines suggest that people who are employed, who have been using for a short time or are under supervision are best candidates for naltrexone. Those who really are well motivated and have good support structures—and perhaps a lot to lose—tend to benefit most from naltrexone.
Since compliance is the biggest issue, the question of who is likely to be compliant with XR-NTX has been examined. People less likely to be compliant include those who identify as homeless, injectors or people with co-occurring mental health issues (13). This is clearly not a population-wide solution.
Reservations Around Naltrexone Research
It’s important to note that there have not been robust clinical trials comparing naltrexone to methadone and/or buprenorphine in the US.
A study in Malaysia, comparing buprenorphine to naltrexone, found the differences in effectiveness were so great that the trial had to be stopped when it became obvious that buprenorphine was a far superior intervention—to continue with the naltrexone arm of the study would have been unethical (14). A study in Iran that compared methadone to naltrexone reported:
“In spite of the fact that patients undergoing methadone treatment had more severe symptoms and prognoses regarding their age, duration of drug abuse, and number of treatment attempts, these patients showed better general health and social functioning comparing to patients undergoing naltrexone treatment during the 6-month period of this study.” (15)
Much of the data has been sourced from studies conducted in Russia (16), where the policy framework excludes agonist or partial agonist interventions and the punitive approaches to addiction treatment leave such studies to be constructed on dubious ethical foundations (17).
A confounder to note is that almost all the research on naltrexone was done with concurrent psychosocial interventions taking place (18). As one study concludes: “The improvements in opioid craving, depression, anxiety, and anhedonia among those who remained in treatment and did not relapse are most likely an effect of treatment success and not specifically related to naltrexone since they occurred regardless of medication group”(19).
Pharmacological, Philosophical and Ethical Concerns
The pharmacology of naltrexone is clear: It blocks opioids. But while this may be great for blocking the effects of heroin, this blockade is non-selective, so all opioids are blocked. In the case of medical trauma it would be very difficult to treat pain, and if high enough doses of opioids were given, this could potentially result in overdose. But of greater concern is the blocking of the body’s endogenous opioids. Recent literature has shown this to have an effect on the processing of social stimuli (20), and it is something that needs to be monitored over time.
Philosophically, I am concerned that we might potentially see the creation of this blockade as the resolution of a heroin use disorder. “Addiction” is not merely a function of pharmacology; it’s extremely complex. To simply block the effects of heroin on, for example, a person living on the streets, is potentially more damaging than allowing that person to use heroin when they choose.
I realize that this thinking will offend some people. But the evidence does not lead me to believe that abstinence is a desirable outcome for all heroin users. It may be for some—but for others, continued heroin use is a choice and for still others, a necessity.
The problem is not the heroin use, nor any drug use, but the context that alienates a person to the degree that their identity and vocation is tied to their drug use. The problem is the drivers that make drug use the most attractive solution to cope with daily life. This cannot be addressed by blocking the drug’s effect.
Agonist therapies reduce harms significantly and still give the person the choice to use heroin, although methadone and buprenorphine both out-perform naltrexone when it comes to reduced heroin use (21). Of course, heroin users should have the additional option to use heroin safely: Diacetylmorphine should be available in supervised injection facilities.
And this leads to my ethical concerns around XR-NTX. When you start offering 28-day naltrexone to people in the presence of loved ones—or if you over-sell it to people who are desperate but unaware of the full consequences, or people who are ambivalent about their drug use—there is the very real risk of coercion.
Already, we are seeing it being used for criminal justice offenders (22). In 2013 a teen was ordered to take XR-NTX, and Ohio courts are ordering heroin users to take it. In a paper published in 2006, Caplan argued that mandated XR-NTX is justified because “heroin addicts” are incapable of making autonomous decisions. Prisoners about to be released in Massachusetts are also offered XR-NTX—but only the first injections, and I wonder what the risk is after the first 28 days? A clinical trial intended to provide additional details about XR-NTX’s efficacy and safety profile (known as a phase 4 pilot) provided pre-release prisoners the option of XR-NTX on release for six months claims that completing the course “may reduce opioid use,” although only 37 percent of the cohort completed the course of injections (23).
Why would we even consider naltrexone as a significant option in addressing heroin use disorders? A basic principle in medical ethics is that when you have an intervention that works, you don’t introduce an alternative unless there are significant advantages and/or improved safety. Naltrexone offers none of this. It produces results no better than placebo in randomized samples and it significantly increases risk of overdose on termination.
What’s more, the Phase 4 pilot does not go head-to-head with OST. A further basic principle of medical ethics is that if a known treatment exists, a placebo-controlled trial is not appropriate. The only justification for its ethical approval was that the research took place in Russia, where OST is banned. The only advantage that I can see for XR-NTX is that it satisfies a moral agenda that prescribes that agonist and partial agonist therapies are somehow “less than” or immoral.
If policymakers really wanted to address the harm caused by drugs, they would first look at the sickening and deadly prohibitionist policies, legal frameworks and agendas that put every drug user at risk. Then, if we are really to address this “heroin epidemic,” we need to increase access to the interventions we know save lives so that people find the time and space to resolve their addictions and stay safe in the interim.
Methadone and buprenorphine need to be accessible to all heroin users with as few restrictions as possible. If we deliver these in compassionate, person-centred harm reduction settings, where autonomy is respected and each person is considered the expert in their own life, then we will see results.
Naltrexone for heroin use disorders should perhaps be available for those who really want it and are fully informed. However, if the current attempts to mandate use of XR-NTX to already vulnerable populations, as we are seeing in the criminal justice system, continue, it will be seen by future generations as another example of how prejudices and stigma have yet again resulted in a gross violation of medical ethics and human rights.
(1) World Health Organisation 2004; Maremmani & Gerra 2010; Lawrinson et al. 2008; Connock et al. 2007; Connery 2015
(2) Mattick et al. 2014; Saxon et al. 2013
(3) WHO 2009
(4) SAMHSA 2015
(5) Adi et al. 2007; Kirchmayer et al. 2002; Minozzi et al. 2011; WHO 2009
(6) Minozzi et al. 2011; Johansson et al. 2006
(7) Johansson et al. 2006
(8) Minozzi et al. 2011
(9) Sullivan et al. 2013
(10) Dijkstra et al. 2007
(11) Davoli et al. 2007
(12) Gibson & Degenhardt 2007
(13) Cousins et al. 2015
(14) Schottenfeld et al. 2008
(15) Sayyah et al. 2013
(16) Krupitsky et al. 2011; Krupitsky et al. 2015; Krupitsky et al. 2013
(17) Wolfe et al. 2010; Cousins et al. 2015
(18) Meyers et al. 1989
(19) Krupitsky et al. 2015
(20) Wardle et al. 2016
(21) Connery 2015
(22) Lee et al. 2016
(23) Gordon et al. 2015
Shaun Shelly is dedicated to the understanding of drug use, addiction and the development of effective drug policy. He is on the advisory boards of Families for Sensible Drug Policy and Harm reduction Abstinence and Moderation Support Network (HAMS), and is a pioneer of harm reduction in South Africa. His last piece for The Influence was “Did Stephen King, the Master Storyteller, Lose His Truth in Writing About Addiction?” You can follow him on Twitter: @ShaunShelly.