Benzodiazepines
Benzodiazepines are efficacious for treatment of GAD. Several benzodiazepines (see
Table 5.1 for a list of benzodiazepines with an anxiety or anxiety disorder indication) have demonstrated efficacy in the treatment of GAD compared to placebo. Generally, low-potency, long-acting benzodiazepines are considered to be safe and effective. High-potency, short-acting compounds, such as the immediate-release formulation of alprazolam, although effective as well, carry a higher risk of dependence and interdose rebound symptoms, unless an extended-release formulation is used. Typically, dosages of 15 to 30 mg per day of diazepam or the equivalent are effective, but occasional patients have required the equivalent of 40 to 60 mg per day of diazepam. Similarly, daily doses of clonazepam between 0.5 and 2 mg, alprazolam (in its sustained-release formulation) doses between 1.0 and 4 mg, and lorazepam doses between 2 and 8 mg are typically used in the treatment of GAD. Some patients with generalized anxiety improve with short-term treatment (2 to 6 weeks), but the majority will have recurrences if treatment is stopped at that time.
Long-term treatment appears to be safe and effective for many patients, who may continue their medication for years; nonetheless, it is worth trying to taper the medication slowly and periodically to see if the underlying condition
has remitted. When tapering medication, it is important to distinguish recurrence of the original symptoms from
transient rebound or withdrawal symptoms.
The benzodiazepines are a group of closely related compounds. The name benzodiazepine is derived from the fact that the structures are composed of a benzene ring fused to a seven-member diazepine ring. A large number of benzodiazepines (
Table 5.1) are currently available. The benzodiazepines are rarely associated with death from overdose in the absence of other drugs. The effectiveness and relative safety of the benzodiazepines have led to their extensive use in the treatment of GAD. Benzodiazepines should be
avoided in patients with a history of alcohol or drug abuse unless there is a compelling indication, no good alternative, and close follow-up. Most benzodiazepines are well absorbed when given orally on an empty stomach; many achieve peak plasma levels within 1 to 3 hours, although there is a wide range among the benzodiazepine drugs approved for anxiety or anxiety disorders (
Table 5.2). Antacids seriously interfere with benzodiazepine absorption; thus, benzodiazepines should be taken well ahead of any antacid dose. The rate of onset of action after an orally administered benzodiazepine (
Table 5.2) is not an important variable in choosing a drug for the treatment of GAD, as the symptoms are chronic in nature. The available benzodiazepines differ markedly in the rate of onset of their therapeutic effect (
Table 5.2), offering a wide choice of drugs to fit the patient’s needs. For benzodiazepines, simple half-life data are potentially misleading regarding duration of clinical effect. Clinical efficacy depends on the presence of at least a minimum effective concentration in the blood, which is reflected by levels in well-perfused tissues such as those in the brain. After a single dose, the levels may decrease to ineffective concentrations, either by being distributed into peripheral tissues, such as fat (α-phase), or by metabolic inactivation or elimination from the body altogether (β-phase). The volume of distribution represents the size of the pool of tissues into which the drug may be drawn; this is determined by the drug’s lipid solubility and tissue-binding properties. With repeated drug dosing, its volume of distribution becomes saturated, and the elimination half-life becomes the more important parameter in describing its behavior. Benzodiazepines differ markedly in their half-lives of distribution and elimination,
producing varying clinical effects. For example, the distribution half-life (α-phase) of oral diazepam is 2.5 hours, whereas the elimination half-life (β-phase) is more than 30 hours. Desmethyldiazepam, diazepam’s major active metabolite, extends the overall elimination half-life to 60 to 100 hours (up to 200 hours in elderly patients). This means that a single dose of diazepam will be active for a relatively short period based on the rapid distribution of the drug, whereas with repeated administration,
elimination half-life becomes the important parameter to consider, making diazepam a very long-acting drug (i.e., one that will accumulate in the body to high levels). Conversely, despite the relatively short elimination half-life of lorazepam (10 hours) and its lack of active metabolites, it has a smaller volume of distribution than diazepam and therefore a longer action when given as a single dose. Another important clinical issue related to duration of effect applies to the use of the high-potency, short-acting benzodiazepines, immediate-release alprazolam and, to a lesser extent, lorazepam. Such drugs pose a potential clinical problem because their high potency may make them more liable to cause dependence and their rapid termination of effect unmasks any dependence that develops. Patients may therefore experience rebound symptoms. Such problems can be addressed by switching to longer-acting drugs when indicated (e.g., replacing alprazolam with clonazepam).
Metabolism. Except for lorazepam and oxazepam, the commonly used benzodiazepines in the treatment of GAD are metabolized by hepatic microsomal enzymes to form demethylated, hydroxylated, and other oxidized products that are pharmacologically active. Most of the benzodiazepines are substrates of the cytochrome P450 3A4; therefore, they tend to accumulate when coadministered with inhibitors of such pathways. The active metabolites of benzodiazepines are, in turn, conjugated with glucuronic acid; the resulting glucuronides are inactive, and, because they are more water soluble than the parent compounds, they are readily excreted in the urine. Some of the active metabolites of benzodiazepines, such as desmethyldiazepam, have extremely long half-lives and with repeat dosing may come to represent most of the pharmacologically active compound in serum. In contrast, under normal circumstances (e.g., excluding cirrhosis) the active metabolic products of alprazolam are of little clinical importance. Lorazepam and oxazepam are metabolized only by conjugation with glucuronic acid with no intermediate steps, and they have no active metabolites. Unlike the pathways involved in the initial metabolism of other benzodiazepines, glucuronidation is less affected by aging and liver disease; thus, if benzodiazepines are to be used in elderly patients or those with cirrhosis, lorazepam and oxazepam are the drugs of choice. In liver cirrhosis, the elimination of benzodiazepines metabolized by oxidation and demethylation may be reduced by as much as fivefold; thus, routine doses could lead to toxicity. In cirrhosis, even alprazolam may accumulate to dangerous levels.
Mechanism of Action. Acting through its γ-aminobutyric acid A (GABA
A) receptor, the amino acid neurotransmitter GABA is the major inhibitory neurotransmitter in the brain. GABA
A receptors are ligand-gated channels, meaning that the neurotransmitter-binding site and an effector ion channel are part of the same macromolecular complex. Because GABA
A receptor channels selectively admit the anion chloride into neurons, activation of GABA
A receptors hyperpolarizes neurons and thus is inhibitory on neuronal firing.
Benzodiazepines produce their effects by binding to a specific site on the GABAA receptor. The pharmacology of GABA
A receptors is complex; GABA
A receptors are the primary site of action not only of benzodiazepines but also of barbiturates and of some of the intoxicating effects of ethanol. GABA
A receptors comprise multiple subunits. GABA
A/benzodiazepine
receptors have a pentameric structure made up of a coassembly of subunits. There are six known α, four β, three γ, two p, and one δ subunits. GABA
A receptors in the brain are most commonly composed of different subtypes of α, β, and γ subunits, with different subunit composition in different brain regions. Series of transgenic mice have been constructed in which each α-subunit was genetically disrupted by a point mutation. This selective inactivation strategy revealed that GABA
A receptors containing the α1 subunit mediate sedation, anterograde amnesia, and part of the seizure protection, whereas GABA
A receptors containing α2 subunits are thought to mediate the anxiolytic effects of benzodiazepines. This finding has led to attempts to develop drugs that selectively interact with different GABA
A receptor subtypes. Benzodiazepines, barbiturates, and alcohol allosterically regulate the GABA
A receptor (changing its conformation) so that it has a greater affinity for its neurotransmitter GABA. At higher doses, barbiturates and ethanol, but not benzodiazepines, can also open the chloride channel within the receptor independent of GABA. The fact that benzodiazepines, barbiturates, and ethanol all have related actions on a common receptor explains their pharmacologic synergy (and therefore the dangers of combined overdose) and their cross-tolerance. Their cross-tolerance is exploited in detoxification of alcoholics with benzodiazepines.
Clinical Uses of Benzodiazepines in GAD. Most patients do not misuse or become addicted to benzodiazepines, though a prior history of alcohol or other substance abuse is a relative contraindication to the use of benzodiazepines. Patients with a history of substance abuse or dependence may become addicted to benzodiazepines. Clinically, the major pharmacologic problem with benzodiazepines is their tendency to cause physiologic dependence, that is, a risk of significant discontinuation symptoms, especially with long-term use. Discontinuation symptoms may pose a serious clinical problem in some patients, causing worsening of (rebound) insomnia, distress, or inability to discontinue treatment. Prior to starting a benzodiazepine, patients should be cautioned about possible sedation and warned not to drive vehicles or operate dangerous machinery until it is determined that the dose does not affect performance. Patients should be instructed to take their medication on an empty stomach and not concomitantly with antacids because meals and antacids may decrease absorption.
When starting benzodiazepines for the treatment of GAD, lower doses (e.g., diazepam, 2 to 5 mg three times daily) should be used initially to assess patient sensitivity to the drug and to avoid initial oversedation. The dosage can be slowly increased until a therapeutic effect occurs. Dosage titration with long-acting drugs (e.g., diazepam, chlordiazepoxide, or clorazepate) should proceed more slowly because the drugs reach steady-state levels over a period of several days. Dosages of short-acting drugs (e.g., lorazepam or alprazolam) can be increased more rapidly (e.g., after 2 days).
In follow-up, patients should be asked not only about efficacy but also about side effects. Patients who complain of excessive sedation may do better with a temporary dosage reduction; over time, most individuals develop tolerance to sedative effects. Patients on short-acting compounds such as alprazolam should be questioned about interdose rebound anxiety, which can be addressed by increasing dosing frequency or using the extended-release formulation. Intolerance to or lack of efficacy of a benzodiazepine may result from pharmacokinetic factors, as with sedation or interdose rebound, and thus may improve with the switch to an agent with a different profile (e.g., alprazolam to clonazepam). An alternative approach is the switch to another class of drugs (e.g., an antidepressant).
Tolerance and Discontinuation Symptoms with Benzodiazepines. The benzodiazepines may induce physiologic dependence. Physiologic dependence must be distinguished
from addiction, which is defined as compulsive, out of control use of a drug despite negative consequences. Addiction to benzodiazepines in nonsubstance abusers may occur, but is rare. Dependence, which occurs with many classes of both psychotropic and nonpsychotropic medications, represents an adapted state of the body to the drug, such that symptoms emerge on tapering of doses, discontinuation of the drug, or even between doses. Discontinuation symptoms can be conceptually divided into (a) recurrence of the original disorder, (b) rebound (a marked temporary return of original symptoms), and (c) withdrawal (recurrence of the original symptoms plus new symptoms, which for benzodiazepines might include tachycardia or elevations in blood pressure). In clinical practice, these syndromes demonstrate a great deal of overlap and frequently coexist. The nature of the symptoms and their time course may help in making distinctions.
Recurrences reflect the loss of therapeutic benefit and typically do not subside with time; the symptoms are generally indistinguishable from those present prior to treatment. Generally, the response to recurrence of the original disorder is resumption of therapy.
Rebound symptoms occur soon after discontinuation and generally represent a return of original symptoms, such as anxiety or insomnia, but at a greater intensity than the original symptoms. The response to rebound symptoms is to observe whether they resolve quickly or to resume therapy and then taper the benzodiazepine more slowly. For some high-potency compounds with a short half-life, such as alprazolam, rebound symptoms occasionally occur even during maintenance therapy as blood levels between doses reach their nadir. If interdose rebound symptoms or rebound occurring with attempts to decrease the dosage represent a serious clinical problem, a switch to a compound with a longer half-life may prove helpful.
The onset of withdrawal symptoms generally reflects the half-life of the drug used, usually 1 to 2 days after the last dose for short-acting drugs, 2 to 5 days for long-acting drugs (although symptoms beginning as late as 7 to 10 days have been reported). Withdrawal symptoms generally peak days after onset and slowly disappear over 1 to 3 weeks. In contrast to recurrence and rebound, withdrawal syndromes include symptoms that the patient has not previously experienced. Benzodiazepine withdrawal symptoms include anxiety, irritability, insomnia, tremulousness, sweating, anorexia, nausea, diarrhea, abdominal discomfort, lethargy, fatigue, tachycardia, systolic hypertension, delirium, and seizures.
The risk of developing dependence and thus rebound and withdrawal symptoms is higher with long-term treatment, higher doses, and higher-potency drugs. The likelihood and severity of rebound and withdrawal symptoms also reflect the half-life of the compound; such symptoms occur more frequently and are generally more severe with compounds with a short half-life.
Although dependence can be induced by any benzodiazepine, rebound and withdrawal symptoms are relatively uncommon with low-potency and long-acting drugs and are typically mild and self-limited when they do occur. Dependence is most likely, and rebound and withdrawal symptoms are most severe with short-acting benzodiazepines, such as alprazolam and lorazepam. These are the compounds that are most likely to produce delirium and seizures after abrupt discontinuation from high doses. In addition to the more common benzodiazepine withdrawal symptoms, severe dysphoria and psychotic-like symptoms have been reported in patients discontinuing alprazolam. Because they are atypical, these symptoms may be extremely confusing on presentation unless a history of alprazolam discontinuation is obtained.
For some patients, discontinuation of alprazolam may be easier via a switch to equipotent doses of a longer-acting, high-potency benzodiazepine such as clonazepam (see next section on switching from alprazolam to clonazepam). A small number of reports also have suggested that anticonvulsants such as carbamazepine,
valproic acid, and gabapentin (Neurontin) may be useful adjuncts for alprazolam withdrawal. Although this strategy may aid an individual patient, a large controlled study failed to show evidence of efficacy for carbamazepine.
Switching from Alprazolam to Clonazepam. Despite the apparently equivalent efficacy of alprazolam and clonazepam for anxiety disorders and the recent introduction of the extended-release formulation of alprazolam, which has greatly reduced the need for a switch, there are still clinical circumstances in which it is helpful to switch patients from alprazolam to clonazepam. These circumstances include signi ficant interdose rebound anxiety or difficulty with tapering and discontinuing alprazolam. As described previously, these issues reflect the high potency and short half-life of alprazolam. Switching to clonazepam appears to address these clinical problems because clonazepam is potent enough to replace alprazolam but has a long half-life (1 to 2 days). One method is based on an open study of patients with panic disorder. The switch takes approximately 1 week (the minimum time to reach a steady-state level of clonazepam).
1. Clonazepam is given at half the total daily alprazolam dose, divided into an early morning and a midafternoon dose.
2. Regular alprazolam doses are stopped but, during the first 7 days, alprazolam can be taken as needed up to the full amount taken previously.
3. Alprazolam is stopped entirely after day 7.
4. If more medication is needed after day 7, clonazepam is increased by 0.25 to 0.5 mg every week until efficacy is reestablished.
Benzodiazepine Abuse. In contrast to public impressions, it appears that few patients who have received benzodiazepines for valid indications become abusers (i.e., increase their dosage without medical supervision and take the drugs for nonmedical purposes) or addicted in the sense of using compulsively. Most abusers of benzodiazepines also have abused other drugs. Serious abusers of CNS depressants may use the equivalent of hundreds of milligrams of diazepam per day. Serious CNS depressant abusers should be detoxified as inpatients using either phenobarbital or a long-acting benzodiazepine, such as diazepam, as the detoxification agent.
Benzodiazepine Use in Elderly Patients. Slowed hepatic metabolism and increased pharmacodynamic sensitivity mean that great care must be taken when prescribing benzodiazepines in elderly patients. In general, short-acting benzodiazepines are safest, especially those metabolized by glucuronidation alone (lorazepam and oxazepam). In one study of patients older than 65 years, use of benzodiazepines with an elimination half-life of more than 24 hours, but not benzodiazepines with a short half-life, was associated with a 70% increase in the risk of hip fracture due to falls compared with individuals not using any psychotropic drugs. Accumulation of long-acting benzodiazepines must always be considered in the differential diagnosis of delirium or rapid cognitive decline in elderly patients.
Benzodiazepine Use in Pregnancy. Earlier reports associating diazepam with both cleft lip and cleft palate have not been substantiated. A cohort study and some but not all case-control studies suggest that benzodiazepines may be safe during pregnancy. However, it would be wise to avoid benzodiazepines, especially early in pregnancy, unless there are compelling reasons for their use.
Benzodiazepine Side Effects and Toxicity. Fatigue and drowsiness are the most common side effects associated with benzodiazepine treatment. In addition, impairment of memory and other cognitive functions and impairments of motor coordination may occur. The benzodiazepines have little effect on autonomic function.
Thus, adverse effects on blood pressure, pulse, and cardiac rhythm are not typically seen. The development of these side effects depends on dosage used (concomitant use of other medications, especially CNS depressants, and alcohol) and the sensitivity of the individual being treated. With repeated dosing, most patients develop tolerance to sedation. The suggestion that automobile accidents are more likely to occur among benzodiazepine users (assuming tolerance to the early sedative effects) is complicated by the possibility that the condition being treated (e.g., anxiety, insomnia) may be a contributing factor. The interpretation of laboratory studies of attention, cognitive control, and driving ability are difficult to generalize to real-life situations.
Acute dosages of benzodiazepines may produce transient anterograde amnesia. This effect appears to be independent of sedation; acquisition of new information is specifically impaired. The risk of anterograde amnesia appears to be worsened by concomitant ingestion of alcohol.
Uncommon side effects include dysarthria, confusion, abnormal coordination, ataxia, depression or worsening of mood (see below), dry mouth, constipation, nausea, slurred speech, dizziness, and tremor. Side effects due to rapid decrease or abrupt withdrawal from benzodiazepines may include agitation, heightened sensory perception, paresthesias, muscle cramps, muscle twitching, diarrhea, reduced concentration, worsening of mood, anxiety, nervousness, restlessness, sleeping difficulties, insomnia, tremors, and, in rare cases, seizures and hallucinations.
Benzodiazepine-Induced Disinhibition. Reports of paradoxical reactions to benzodiazepines (disinhibition), in particular describing rage outbursts or aggression in patients on chlordiazepoxide, diazepam, alprazolam, or clonazepam, have been published. Disinhibition can probably occur with any benzodiazepine, but the lower potency, slowly absorbed oxazepam may be less likely to trigger this effect. Many clinicians feel that the highest incidence of disinhibition occurs in personality disorder patients with prior histories of dyscontrol. When paradoxical excitement occurs in a patient given a benzodiazepine in an emergency department or inpatient ward, the administration of an antipsychotic drug is often effective in reversing the state.
Benzodiazepine-Induced Depression. All benzodiazepines have been associated with the emergence or worsening of depression; whether they were causative or only failed to prevent the depression is unknown. If the depression occurs during the course of treatment, the benzodiazepine can be combined with or replaced by an antidepressant.
Benzodiazepine Overdose. With respect to lethality, benzodiazepines have proved to be relatively safe in overdose in that benzodiazepines alone have only rarely been implicated in fatal overdoses. However, when combined with other CNS depressants, such as alcohol, barbiturates, or narcotics, benzodiazepines may contribute to the lethality of the overdose.
The treatment of benzodiazepine overdose includes induction of emesis or gastric lavage, when appropriate, and supportive care for patients who are stuporous or comatose. The benzodiazepine antagonist flumazenil is available for the treatment of benzodiazepine overdose. In benzodiazepine-dependent patients, this drug may precipitate withdrawal symptoms in analogy with the actions of naloxone in opiate-dependent individuals.
Benzodiazepine Interactions with Alcohol and Other Drugs. Serious pharmacokinetic drug interactions are rare with benzodiazepines but may occur (
Table 5.3). Benzodiazepines can cause a mild to moderate increase in CNS depression caused by
coingested alcohol; when taken together in overdose, ethanol and benzodiazepines can result in death.
Buspirone
An alternative to antidepressants or benzodiazepines for GAD is buspirone (BuS-par), a partial agonist of the serotonin 5-HT1A receptor. Buspirone has also slight affinity for the dopamine D2 receptors. Buspirone may be a good initial choice in patients with GAD who are at elevated risk for benzodiazepine abuse, but it is not indicated in GAD with comorbid depression and/or panic disorder, where antidepressant monotherapy is typically the first choice. Buspirone, which has been approved for treatment of GAD, is a member of a chemical group called the azaspirodecanediones. Buspirone has no direct effects on GABAA receptors, it has no pharmacologic cross-reactivity with benzodiazepines or barbiturates, and it lacks the sedative, anticonvulsant, and muscle relaxant effects of benzodiazepines. A major advantage of buspirone is that it does not produce dependence, and it has no abuse potential.
Buspirone is believed to exert its anxiolytic effect by acting as a partial agonist at 5-HT1A receptors. Because 5-HT1A receptors are autoreceptors, their activation by buspirone decreases serotonin release. Investigations of the role of serotonin in anxiety suggest that there is a complex relationship between the neurotransmitter and anxiety symptoms. This complexity is highlighted by the fact that SSRIs increase synaptic serotonin rather than decrease it. Ultimately because both buspirone and SSRIs produce therapeutic actions only after a latency of days to weeks, it is likely that their beneficial effects reflect adaptations within certain circuits to drug stimulation. Of interest, buspirone has an active metabolite, 1-phenylpiperazine, that acts via α2-adrenergic receptors initially to increase the rate of firing of locus coeruleus neurons. Because the ratio of the plasma level of this metabolite to the parent compound increases with treatment over time, this stimulation of adrenergic systems may play a key role in the agent’s therapeutic as well as adverse effects.
Buspirone is 100% absorbed from the human gastrointestinal tract but undergoes extensive first-pass metabolism by the liver so that only 4% may be bioavailable. Buspirone is metabolized by the liver; the half-life of the parent compound is 2 to 11 hours. Oxidative metabolism of buspirone is carried out through
cytochrome P450 3A4. Therefore, inhibitors of 3A4 activity such as grapefruit juice may significantly elevate buspirone blood levels.
When used to treat GAD, buspirone has been found to be as effective as standard benzodiazepine treatments in some but not all studies. Buspirone is most often ineffective as a sole treatment for panic disorder. It appears that patients with GAD who have taken benzodiazepines within 4 weeks prior to taking buspirone may be less likely to benefit from buspirone. Unlike benzodiazepines, buspirone is effective only when taken regularly. It takes 1 to 2 weeks to show its initial effects, and maximal effectiveness may be reached only after 4 to 6 weeks. This must be clearly explained to patients who are accustomed to using benzodiazepines. Because of this time course of effectiveness, buspirone is not useful in emergencies or when rapid onset of anxiolysis is required. The initial dosage of buspirone is 5 mg three times daily; in most trials, 20 to 30 mg per day in two or three divided doses has been effective, but a total dosage of up to 60 mg per day may be required for an optimal response. Buspirone is available in 5-, 10-, 15-, and 30-mg tablets.
Because of its lack of cross-reactivity with benzodiazepines, buspirone cannot prevent benzodiazepine withdrawal symptoms. Therefore, when switching patients from a benzodiazepine to buspirone, the benzodiazepine must be slowly tapered as if no new drugs were being introduced. If buspirone is started before the taper has concluded (which should be safe), it is important not to confuse benzodiazepine withdrawal or rebound symptoms with buspirone side effects.
Buspirone does not cause sedation; however, it may occasionally produce restlessness. It does not appear to impair psychomotor performance. The other most common side effects of buspirone are dizziness, headache, light-headedness, gastrointestinal distress, nausea, insomnia, paresthesia, and drowsiness. Buspirone does not appear to be highly toxic in overdose. Buspirone has been somewhat disappointing in general clinical use in that a smaller percentage of anxious patients benefit from buspirone than from benzodiazepines. Whether this reflects inappropriate expectations and inadequate dose and duration of treatment on the part of both physicians and patients who are accustomed to the rapid effects of benzodiazepines is unclear. For patients who respond to buspirone, it has the marked advantages of being free of sedation, lacking any prominent discontinuation symptomatology.