Abstract
See Review by Moylan et al., 2012, 46(3): 212–224
Benzodiazepines have recently turned 50, as chlordiazepoxide, the first member of the benzodiazepine family, was introduced into clinical practice in 1960. During the last half a century, these medications have experienced both stellar moments and castigation, provoking strong feelings and dividing the medical profession. When it comes to the anxiety disorders, treatment guidelines now regard benzodiazepines as a second- or third-line choice of medication and perhaps suitable only for short-term treatment or in some treatment-resistant cases. While it may seem that the role of benzodiazepines has thus been circumscribed, a single conversation with almost any medical practitioner would suggest that this is not the case and that benzodiazepine controversy is alive and well.
In this issue of the journal, Moylan et al. (2012) examine the role of alprazolam in the treatment of panic disorder (PD) in Australia and conclude that it “merits renewed consideration”. They arrive at this conclusion after discussing various aspects of alprazolam use and their article calls for further scrutiny of the underlying issues.
Why are benzodiazepines still popular in the treatment of panic disorder?
Moylan et al. (2012) note that despite guideline recommendations, there has recently been a significant increase in the use of alprazolam in Australia. They offer several explanations for this. More broadly, alprazolam and benzodiazepines in general remain popular in the treatment of PD for several reasons:
consistent effectiveness for relieving various physical symptoms of anxiety, tension and sleep disturbance, which is particularly relevant for PD individuals because they often ask for quick alleviation of anxiety and distress;
quick onset of action;
generally good tolerability, with side effects being an uncommon reason for discontinuation;
possibility of an ‘as-needed’ (prn) administration, which may be useful to some PD patients, especially before they have to face panic-triggering situations;
relative safety in overdose, if not taken in combination with other agents;
some disappointment with antidepressants in the treatment of PD, because of their slow onset of therapeutic action, unpredictable and occasionally troublesome side effects and inconsistent or unreliable effectiveness. Evidence is accumulating regarding problems associated with long-term use of the selective serotonin reuptake inhibitors (SSRIs) (Moret et al., 2009), which are now considered first-line pharmacotherapy for PD. A shift from benzodiazepines to SSRIs as the preferred treatment for anxiety disorders has also been questioned on the grounds that adequate comparative evidence of efficacy is lacking (Berney et al., 2008) and that dependence issues related to SSRIs differ very little from those related to benzodiazepines (Nielsen et al., in press).
Should a notion that benzodiazepines are addictive be endorsed?
It is often assumed that benzodiazepines are addictive, but the basis for this notion remains unclear. This is largely related to various conceptualisations of addiction. Moylan et al. (2012) have chosen a broad, behaviour-based definition of addiction that does not necessarily include tolerance. Likewise, the recent American Society of Addiction Medicine definition of addiction refers to “impairment in behavioural control, craving and diminished recognition of significant problems with one’s behaviour and interpersonal relationships” (American Society of Addiction Medicine, 2011).
Even if one were to espouse such definitions, would benzodiazepine dependence qualify as an addiction? In the absence of other drug misuse, patients with anxiety disorders treated with benzodiazepines very rarely exhibit an all-encompassing preoccupation with benzodiazepines, craving for these medications, uncontrollable benzodiazepine-seeking behaviour (including a determination to obtain them by almost any means) and damage to their reputation, resulting in financial and/or legal problems due to benzodiazepine use (Andersch and Hetta, 2003; Lader, 2011). Therefore, if benzodiazepines are used to treat an anxiety disorder, addiction does not seem to be an adequate term to describe dependence-producing effects of these agents. Unfortunately, alternative terms such as therapeutic or non-addictive dependence (Starcevic, 2007) or low-dose iatrogenic dependence (Lader, 2011) have not been widely accepted.
Without sufficient understanding of the underlying mechanisms and consensus on terminology, we can at least avoid the term ‘addiction’ in the context of therapeutic benzodiazepine use. Making a distinction between dependence associated with benzodiazepine use and addiction to substances like alcohol or heroin is not a pedantic pursuit. It will help prevent further confusion and rift, especially when explaining to patients the effects of benzodiazepines. For example, it is important to clarify that benzodiazepine dependence and addiction per se only share the occurrence of withdrawal symptoms upon abrupt discontinuation, so that when a patient is ready for benzodiazepine cessation, it needs to be done through a gradual dose reduction to prevent withdrawal symptoms.
Should other benzodiazepines replace alprazolam in the treatment of panic disorder?
Moylan et al. (2012) state that alprazolam has no advantage over other benzodiazepines in the treatment of PD based on their meta-analysis of the efficacy of alprazolam, lorazepam, diazepam and clonazepam, and on their suggestions that in comparison with other benzodiazepines, alprazolam is more toxic and therefore more dangerous in an overdose, more likely to be abused and more difficult to discontinue. However, except for clonazepam, there is more evidence of PD efficacy for alprazolam than for other benzodiazepines, and results of one meta-analysis, with all the limitations inherent to this method, should not be considered definitive. A finding that alprazolam is more toxic than other benzodiazepines needs to be replicated; benzodiazepines are generally safe in an overdose situation and their use is usually not precluded by a high suicide risk. Likewise, alprazolam abuse is not of great relevance in the setting of PD, unless there is a co-occurring substance use disorder. Also, some evidence suggests that alprazolam may have a lower abuse potential than diazepam (Pradel et al., 2010). Although it may appear that short-acting benzodiazepines (e.g. alprazolam or lorazepam) are more difficult to discontinue than the longer-acting ones (e.g. diazepam or clonazepam), this has not been systematically investigated. Moreover, short-acting benzodiazepines may be more suitable for some PD individuals and for prn use. Inter-dose anxiety with alprazolam may be addressed by its more frequent, tid or qid dosing. All this suggests that it is premature to replace alprazolam with other benzodiazepines in the treatment of PD. What we need are controlled studies in PD patients that would compare efficacy, tolerability and dependence-related issues between various benzodiazepines.
Future role of benzodiazepines
It may indeed be the time to reconsider the role of benzodiazepines in the treatment of anxiety disorders. While in an ideal world they would not need to be used, in the reality of clinical practice there will always be people with anxiety disorders who do not get better with other medications or psychological treatment or who cannot use these therapeutic modalities, thereby qualifying for a trial with a benzodiazepine. There are also patients for whom benzodiazepines may be suitable as first-line treatment. Rather than imposing restrictions on prescribing benzodiazepines, indications for their use need to be carefully reviewed. We should not feel embarrassed that benzodiazepines are a part of our therapeutic armamentarium. Instead, we would do well to make the most of them and promote their sensible and safe use.