Abstract
There has been recent heightened concern over heroin use in our community. The media has portrayed an ‘epidemic’, the rate of fatal overdose is rising, and governments are seeking various solutions, including the establishment of supervised injecting facilities, new treatment approaches and improved prevention programmes. Evidence indicates that there has been an increase in heroin use and associated harms in Australia [1]. The number of deaths attributed to heroin and other opiate overdose among Australian adults aged 15–44 years increased from 6 in 1964 to 600 in 1997 [1]. Data on ambulance attendance at non-fatal heroin overdoses in Victoria demonstrates a steady rise from a monthly rate of 157 in June 1998 to 379 in January, 2000 [2, 3]. In addition, greater numbers of people are seeking treatment for heroin dependence [4]. In spite of these alarming figures, it should be pointed out that estimates of the actual number of regular, dependent heroin users in Australia remains relatively low. Most recent estimates indicate that approximately 74 000 people are heroin dependent, with a prevalence rate of 6.9 per 1000 adults aged 15–54 years [1]. Thus, a relatively small number contribute to a substantial public health problem.
Treatment approaches for heroin dependence include: drug withdrawal; abstinence based programmes (for example, therapeutic communities); and a variety of pharmacotherapies. Drug withdrawal is concerned with safe and comfortable neuro-adaptation reversal. Heroin withdrawal is rarely life-threatening and treatment usually takes 5–7 days. However, the greatest challenge for most patients is not undergoing the withdrawal per se, but developing and sustaining a drug-free lifestyle. Hence, postwithdrawal treatment options are crucial in maintaining behaviour change. These postwithdrawal abstinence-based programmes, such as therapeutic communities, aim to develop effective life-coping skills for patients. Maintenance pharmacotherapy is an alternative option. The rationale for maintenance pharmacotherapy is that the provision of a legal, stable source of opiate, in conjunction with appropriate counselling and support services, enables the individual to cease heroin use, and the associated behaviours, such as drug-seeking. During the period of maintenance on a pharmacotherapy, other lifestyle changes can be achieved, such as re-establishing relationships, gaining employment, and making meaningful contributions to the community. The most efficacious form of treatment for heroin dependence is methadone maintenance, a substitution pharmacotherapy programme [5].
Until recently, these were the only three options available to heroin users seeking treatment in Australia. Pleasingly, however, some new treatment options are becoming available. These are all pharmacotherapies: new maintenance treatment options such as buprenorphine and Leevo-alpha acetyl methadol (LAAM), new withdrawal treatment options such as buprenorphine; and new relapse prevention options, such as naltrexone.
Australia has experienced a surge in interest in naltrexone, with the drug's registration in Australia in 1999 for the treatment of both alcohol and heroin dependence. The media has given naltrexone a high profile and raised unrealistic expectations of the value of naltrexone in treating heroin dependence. Given the high prevalence of depression in heroin users, psychiatrists may find themselves called upon to either prescribe naltrexone to their patients or to treat depressed opiate-dependent patients who are in receipt of naltrexone.
Naltrexone treatment for heroin dependence
Naltrexone is an opiate antagonist which binds at the opioid receptors and blocks the effects of exogenous opioids. Originally developed as a relapse prevention agent, problems were experienced with the induction of patients onto naltrexone: naltrexone precipitates withdrawal in individuals who still have opioids present, hence it is recommended that induction onto naltrexone not occur until at least 5–7 days after last use of opioids. This in turn results in few patients entering naltrexone treatment (as many are not able to achieve this period of abstinence). In response, practitioners have examined the capacity to use naltrexone deliberately to precipitate a withdrawal, and lead to induction onto naltrexone for relapse prevention. These ‘rapid’ and ‘ultra rapid’ withdrawal techniques using naltrexone have yet to be demonstrated to be safe and efficacious [6].
The indicated use of naltrexone is as a postwithdrawal relapse prevention intervention. Clinical studies have indicated that a 50-mg dose blocks the effects of 25 mg of intravenously administered heroin for longer than 24 h [7]. Doubling the dose to 100 mg lengthens the period of blockade to 48 h [8]. Naltrexone has a good side-effect profile, does not produce tolerance or dependence and unlike disulfiram (Antabuse), naltrexone is not an aversive therapy.
It is important to understand the clinical rationale for naltrexone effectiveness. There are three components to naltrexone's effect upon heroin-seeking behaviour: conditioning theory, delaying reinforcement, and reduced cravings. Conditioning plays a large role in the initiation and continuation of drug use, with the euphoric effects of opiates acting as strong positive reinforcement for further use. If the opiate-dependent individual is exposed to drug-related cues without the positive reinforcement of euphoria, over time, drug-seeking behaviour and craving is extinguished. Secondly, naltrexone's half-life means that patients cannot experience the euphoria of heroin until at least 24 h after the last dose. This means that they are forced to delay any drug-seeking behaviour– resulting in sufficient time to review their decision and perhaps seek additional support. Lastly, naltrexone has been reported to directly reduce craving for heroin. It is unclear precisely how this effect is achieved, and it appears to vary quite considerably between individuals [9–11].
In spite of the length of naltrexone availability, particularly in the USA, there have been only a handful of randomized controlled trials on the efficacy of naltrexone maintenance with heroin dependent patients. A recent Cochrane Review [12] concluded that there is to date insufficient evidence to evaluate the efficacy of naltrexone for heroin dependence; however, in select patient groups where motivation is high, naltrexone outcomes are superior to controls. This equivocal research evidence base goes some way to explaining the recent Commonwealth Government's decision not to subsidize naltrexone on the Pharmaceutical Benefits Scheme. In spite of the less than ideal efficacy data, naltrexone does provide an alternative treatment option for some people, and hence perhaps some new hope in dealing with a heroin dependency problem.
This paper is concerned with examining more closely two clinical aspects of naltrexone treatment: the relationship of naltrexone to depression, and the risk of overdose subsequent to naltrexone treatment.
Depression and naltrexone
Within mental health populations, the rate of heroin use is higher than in the total population. Likewise, within heroin using samples, the rates of mental health problems are high. Recent estimates indicated a lifetime prevalence of depression among heroin users of 41% [13]. Thiry per cent reported a current episode of depression [13].
Various aetiologies may be at work. Some people may have a pre-existing depressive illness, and use drugs, including heroin, as a way of managing their psychiatric illness. Alternately, depression may be secondary or a consequence of drug use. The drug-using lifestyle is stressful, many users are disconnected from their families and the community, and most have poor economic circumstances. Experiencing depression under these circumstances is expected. In addition, treatment for heroin dependence is frequently accompanied by feelings of depression. Withdrawal from heroin is associated with depression and the struggle to remain drug-free coupled with the loss of the drug can lead to feelings of hopelessness and depression. Two further explanations are also possible: the high rate of depression among heroin users may occur due to common risk factors, such as socioeconomic status, or it may co-occur by chance.
While the aetiology may be of academic interest, there are treatment implications. For those with a pre-existing depressive illness, active treatment of the depression, particularly concurrent with drug treatment is important. Likewise those developing depression during the course of drug treatment require active interventions.
One of the reported side-effects of naltrexone is dysphoria [6, 10]. Given the above analysis, however, it is difficult if not impossible to tease out the relationship between depression as a side-effect of naltrexone and depression occurring as part of the course of the treatment response, independent of the drug naltrexone.
Why might there be a relationship between depression and naltrexone? Naltrexone blocks the opioid receptors, and there has been some suggestion that it effectively blocks both exogenous and endogenous opioids. Thus, there is a potential pharmacological basis for naltrexone's depressogenic properties. The existing studies of the relationship between naltrexone and depressive symptoms are extremely limited. Four different samples have been studied: healthy volunteers; ex-heroin users; alcohol dependent samples; and heroin dependent samples. In healthy volunteers, Hollister and colleagues [14] administered a single dose of naltrexone in one study; and a 3-week induction regime in another. In both designs, participants experienced more depression in the naltrexone compared to the placebo conditions. Likewise, other research has found some evidence that in healthy volunteers naltrexone produces or exacerbates depression [15].
It is unclear, however, whether findings from non-drug users can translate to the drug-using population. This is particularly the case where drug users have an already significantly altered receptor system. In an attempt to overcome this, Crowley studied eight ex-heroin users [16]. In spite of significant problems with the study, five of the eight dropped out due to side-effects. Of the three who remained in the study, two were significantly more depressed on naltrexone than placebo.
In alcohol-dependent samples, no significantly greater rates of depression were found in the naltrexone groups compared to the control groups [17, 18]. In a recent Australian placebo controlled study, Latt, Jurd and Wutzke examined the relationship between depression, treatment for alcohol dependence and naltrexone [19]. They found no evidence that those patients receiving naltrexone experienced more depression than the placebo group.
In heroin dependent samples three different results have been reported: non-significant differences between naltrexone and placebo on depression [20]; significantly higher depression in the naltrexone group [21]; and significant decreases in depression over the course of naltrexone treatment [11, 22, 23].
Based on the existing equivocal evidence it is reasonable to conclude that at this time, there is no demonstrable relationship between naltrexone and depression in heroin dependent patients. Nonetheless, individual patients may well experience increase in dysphoria associated with naltrexone treatment.
Naltrexone and risk of overdose
Naltrexone produces blockade of the opioid receptors, hence when naltrexone is present, an opioid overdose is not possible. The affinity of naltrexone at the receptors is higher than heroin. Heroin dependence is a chronic relapsing condition, and many patients return to heroin use, either on a sporadic basis, or resume regular dependent use. All cessation of opiates causes up-regulation and hence reduction in tolerance to opiates. It is well-known that patients are at heightened risk of overdose following a period of abstinence. However, does the abstinence associated with naltrexone, or other antagonists, increase the risk?
It appears that there is a potential biological mechanism to the facilitation of up-regulation with naltrexone. Treatment with an opiate antagonist may produce functional supersensitivity resulting in increased toxic potency of opiates. Animal studies have demonstrated this up-regulation with naltrexone in rodents [24–26]. One study attempted to replicate this finding in healthy volunteers. Cornish administered baseline morphine, then 14 days of naltrexone, followed by a morphine rechallenge [27]. The results were non-significant – there was no evidence of increased sensitivity to morphine following 2 weeks of naltrexone. To our knowledge, this study has not been replicated with either healthy volunteers or with heroin users.
Other sources of evidence, however, are emerging. Miotti and colleagues reported the first evidence of increased rates of overdose in a sample of naltrexone patients [22]. Thirteen of 81 participants overdosed within a 12-month period of study participation.
At Turning Point, we have been conducting a number of clinical trials of different pharmacotherapies for heroin dependence. The three drugs being researched are LAAM, buprenorphine and naltrexone. The studies are randomized controlled efficacy trials, conducted under the International Good Clinical Research Practice (GCRP) Guidelines, with relevant institutional ethics committee approvals. All the studies have standardized reporting of serious adverse events (SAE), as per the national TGA definitions. Non-fatal overdose is classified as an SAE. While still in data collection mode, the total number of SAE's across the studies is currently 48, with a total of 446 patients. Examination of the rates of reporting of non-fatal overdoses across the studies reveals that the rate of non-fatal overdose in the naltrexone study is presently significantly greater than in the other two studies (see Table 1).
Rates of serious adverse events (SAE's) reported in three clinical trials at Turning Point1
Expressed as a percentage, to date 16.5% of the naltrexone study participants have experienced a non-fatal overdose, compared to 0 and 0.5% in the LAAM and buprenorphine studies. Further, the naltrexone study is a 6-month follow-up study, whereas the LAAM and buprenorphine studies are 12 months, hence the per annum rate of nonfatal heroin overdoses in the naltrexone study is higher.
It is important to note that differential reporting may be at work. The LAAM and buprenorphine studies are being conducted in general practice settings and hence reporting of SAE's is reliant upon the GP, whereas the naltrexone study is being conducted on-site at Turning Point, and the degree of vigilance in relation to reporting non-fatal overdose is higher. However, the significantly higher levels of non-fatal overdose in the naltrexone trial cannot be entirely accounted for by differences in reporting rates.
Examination of fatal overdose (mortality) is also informative. Untreated heroin users have a mortality rate between 2 and 4% per annum [28]. Methadone maintenance patients have an estimated mortality rate of less than 1% per annum [29], consistent with the results reported in Table 1. Mortality rates from published naltrexone outcome studies indicate a higher rate (3–5%) [22, 30]. There have been no fatal overdoses in the Turning Point naltrexone study to date.
It is fair to conclude that naltrexone treatment carries a heightened risk of overdose, and patients should be educated regarding reductions in tolerance.
There may in addition be a suicide factor operating. Overdoses are common among heroin users (60% to 70%) [31, 32], it is unclear what proportion is accidental and intentional. In a study, [33] 200 heroin users in methadone treatment, 69% reported that their most recent overdose was accidental, 26% deliberate. The latter were more depressed at the time of interview, with greater suicidal ideation [33].
Conclusion
Naltrexone, a valuable treatment for alcohol dependence, does not raise the risks of depression or overdose. In the context of limited treatment options for heroin dependence, naltrexone represents an important addition. The relationship between naltrexone and depression remains unclear, partly as a result of the complexity in understanding depression in heroin users. Heroin users have elevated rates of depression, and upon treatment entry they are frequently depressed. This may be exacerbated by treatment, or resolve as heroin use becomes manageable. A medication that blocks the opioid receptors, further confounds the picture.
A serious sequel of naltrexone treatment is reduction in tolerance to opiates. While the mechanism is not well understood, research has only demonstrated an effect in animals, there is mounting evidence of increased risk. Given the high drop-out from naltrexone treatment, caution must be exercised. Effectively delivered counseling should be incorporated into treatment programmes.