Using Benzodiazepines Wisely

Hailed as a breakthrough 40 years ago, benzodiazepines (BZPs) have fewer drawbacks than its predecessors. BZPs have a comfortably wide safety margin in overdose, minimal pharmacokinetic drug interactions, and good tolerability, while maintaining effectiveness for a wide range of conditions. For these reasons, BZPs remain among the most widely prescribed drugs worldwide.

Current uses of BZPs are varied and include but are not limited to insomnia, anxiety disorders (e.g., panic disorder, generalized anxiety, phobias, anxiety associated with obsessive compulsive disorder), mania, for behavior control in psychotic patients, catatonia, neuroleptic-induced side effects including akathisia, muscle spasms, seizures, and alcohol withdrawal.

The drawbacks of BZP use have been extensively investigated and reported on. We propose to review these issues and provide guidance for their appropriate use.

Problems associated with BZP use: an inventory

1. Dependency: after several weeks of use, physical dependency can develop and often becomes apparent when attempting to stop extended periods of use. With abrupt withdrawal, patients experience a rapid return of baseline symptoms (relapse) often with exaggerated intensity (rebound) and often accompanied by new symptoms (withdrawal).
2. Excessive sedation: when used for indications other than insomnia sedation is often a disadvantage and may place vulnerable patients (the elderly) at risk.
3. Memory problems: anterograde memory problems have been documented with all BZPs and are especially problematic with triazolam, a rapid-acting BZP with short duration of effect. It is potentially frightening to the patient and the focus of several lawsuits against physicians in the U.S. who failed to forewarn their patients of this possibility.
4. Disinhibition: may be a particular problem with diazepam and triazolam, but has been described with all agents at an estimated rate of <1%.
5. Ataxia: falls are an all too frequent occurrence, especially in the elderly taking BZPs who are at 1.5-3 times the risk for falls.
6. Drug interactions: clinically meaningful pharmacokinetic interactions involving BZPs are not common, though clinicians should be mindful of the potential for drug concentration elevations of the triazolo-benzodiazepines (triazolam, alprazolam, midazolam) when combined with P450 3A4 enzyme inhibitors (e.g., nefazodone, erythromycin, ketoconazole, HIV protease inhibitors, and even grapefruit juice). Of greater concern is the pharmacodynamic interaction with other drugs that produce CNS effects including sedation and/or unsteadiness (e.g., alcohol, antihypertensives). Patients should be forewarned and monitored for any additive untoward effects of this nature.

Triazolam is of specific concern. We concur with the Committee on Safety of Medicines in the UK that concluded the risks associated with triazolam use outweigh the benefits and that this agent should be avoided, especially in the elderly.

Starting BZP Treatment

When the above concerns have not dissuaded the practitioner from starting a BZP, due consideration of the following is encouraged.

Choose a BZP with an onset of action that matches the patient's circumstances. For acute control of anxiety or agitation rapid onset is required. Lorazepam, diazepam and alprazolam are all relatively quick acting. However, lorazepam is recommended due to its moderate duration of effect and low risk for accumulation. Diazepam's onset is very rapid and for some it is associated with a euphoriant effect which has been linked to its heightened abuse potential. Alprazolam, though quick in onset is also quick in offset and once treatment with alprazolam is established it can be particularly difficult to discontinue. For this reason alone, we suggest avoiding alprazolam altogether.

Consider the duration of effect and the risk for accumulation of active metabolites, especially in the elderly and in individuals with hepatic impairment (e.g., chronic alcoholics). Patients at risk for accumulation should avoid long acting BZPs (e.g., diazepam, flurazepam, nitrazepam, bromazepam). Short to intermediate acting agents without active metabolites (e.g., lorazepam, oxazepam, clonazepam, or temazepam) or single step metabolism (glucuronidation) BZPs (e.g., lorazepam, oxazepam) are preferred.

Another consideration is the availability of different dosage forms and strengths. Lorazepam is the most versatile (PO, SL, IM, IV, and PR) followed by diazepam (PO, IV, PR). The flexibility with lorazepam is particularly attractive for unstable hospitalized patients, but may also be important for debilitated patients at home managed by family members or nurses.

A final consideration, or plea, is the notion of one BZP per patient. Though marketed differentially, clinical differences among the BZPs are insignificant. A single agent should be chosen for a patient based on a match between the patient's symptoms and the pharmacokinetic profile of the drug. For differential effects during the day versus evening and night, the dose should be adjusted, not the drug. If >1 BZP has been prescribed, one BZP should be chosen and the equivalent doses of the other agents substituted.

Stopping BZP Treatment

BZP discontinuation can result in relapse, rebound and/or withdrawal. High potency short-acting BZPs are the most difficult to discontinue (e.g., alprazolam) while those with accumulating active metabolites (diazepam) may have the benefit of an inherent self-taper that reduces the likelihood of a severe withdrawal reaction.

1. experience of early relapse, rebound or withdrawal during past discontinuation attempts;
2. use of short-acting agents;
3. regular treatment for longer than 3-4 months; and
4. daily dosage.

However, it is almost never advisable to stop BZP treatment abruptly when treatment has been ongoing for longer than six weeks and is likely best to taper treatment gradually even after a week or two of regular use.

Incorporation of a few measures can improve your chances of a successful and safe BZP taper. First, the plan should be derived in partnership with the patient adjusting it based on their enthusiasm and fears. Second, the taper schedule should allow for some flexibility as the taper period is often long and rarely goes as originally planned. Often it is the last 25% of the original dose that is the most difficult to reduce. The patient should be informed of potential withdrawal symptoms and educated on what measures to take to manage them. This should include explicit instructions regarding the use of rescue doses. Third, switching to a longer-acting agent should be considered, especially for patients who have tried and failed to reduce their current treatment. Switching can be accomplished by calculating the equivalent dose of the substituting benzodiazepine and starting it all at once over 24 hours. The other agent should be stopped with a few doses allowed over the next few days to manage breakthrough symptoms. Switching alprazolam to a longer acting preparation, such as clonazepam, has been recommended, however cross-tolerance is often incomplete adding to the already difficult discontinuation process with this agent. And fourth, an asymptotic dose reduction is advised. This can be accomplished by reducing the dose by a fixed proportion of the current dose (not the initial dose) at a regular interval, for example decrease by 10-20% weekly.

It is hoped that this recount and guide to proper BZP use will remind clinicians of their proper use and limitations and will lead to improved prescribing practices.

Diane McIntosh, M.D.
David M. Gardner, Pharm.D.

References:

1. Janicak PG, Davis JM, Preskorn SH, Ayd FJ. Principles and practice of psychopharmacotherapy. ed. Baltimore: Williams & Wilkins, 1997.
2. Benzodiazepine dependence, toxicity and abuse: a task force report of the American Psychiatric Association. Washington, DC, APA, 1990.
3. Bernstein JG. Handbook of drug therapy in psychiatry. ed. [city]: Mosby-Year Book, Inc, 1995

This article originally appeared in Atlantic Psychopharmacology (Winter 1998) and is reproduced with permission from the editors, Serdar M. Dursan, MD PhD FRCP(C) and David M. Gardner, PharmD.

You can search for abstracts of the above references by following this link: PubMed

This page last updated November 27, 1999