Chapter 42

Hypnotic Medications: Mechanisms of Action and Pharmacologic Effects

Wallace Mendelson
Chapter
Abstract
Most currently available hypnotics induce sleep by acting at various moieties of the GABAA–
benzodiazepine receptor complex. It is one receptor of the family of ligand-gated receptors,
composed of several glycoprotein subunits, each of which appears in multiple isoforms.
Microinjection studies indicate that major neuroanatomic sites of action of hypnotics from a
variety of pharmacologic classes include the preoptic and other areas of the hypothalamus.
Among the latter are areas centered around the function of other neurotransmitters but which
receive significant GABAergic input. Connections between the preoptic area and the forebrain
and brainstem, and the integrative nature of the anterior hypothalamus, help explain the
mechanism by which sleep—and drugs that affect sleep—interact with a variety of physiologic
systems. A number of hypnotic agents with novel mechanisms of action are currently under
development and are likely to be available clinically in the next few years.

INTRODUCTION
Currently, the clinically-used sedative/hypnotics in the United States include five
benzodiazepines (Valium-like compounds) available in both proprietary and generic forms, three
nonbenzodiazepine gamma-aminobutyric acid (GABA) agonists (zolpidem, zaleplon, and
eszopiclone), and a melatonin receptor agonist (ramelteon) available as prescription agents
(Table 42-1). This is a large market, involving approximately 49 million prescriptions in 2006, a
growth of 53% over the previous 5 years. Sales in 2006 were approximately $3.7 billion. During
the same year, roughly the same number of prescriptions was written for nonhypnotics
prescribed for the purpose of aiding sleep; the largest among these was the antidepressant
trazodone. Although these are large numbers, one could make a case that in the population at
large, prescription hypnotics are taken relatively infrequently by insomniacs. In a survey of over
7000 enrollees in five large health maintenance organizations, patients with insomnia were
divided up into those with only sleep complaints (Level 1) and those who believed that their
sleep difficulties had an adverse effect on their daytime functioning (Level 2). The percentage of
Level 1 insomniacs taking prescription and nonprescription hypnotics was 5.5% and 11.2%,
comparable values for Level 2 insomniacs were 11.6% and 21.4%, respectively. Many patients,
of course, receive other classes of prescription drugs given for purposes of helping sleep. One
study of primary care patients in a health maintenance organization indicated that 13% of
insomniacs who were not considered to have affective disorder were receiving antidepressant
medications.
Many persons also self-medicate with alcohol, a telephone survey of approximately 1000
representative adults in the general population found that 10% had done so in the past year. A
survey in the Detroit area reported that during the past year, 13% had used alcohol to aid sleep,
whereas 18% used medication (either prescription or over-the-counter), and 5% had used both.
Thus, alcohol use to help sleepis common even though it exposes the patient to the risk of
ethanol dependence, and it is also relatively ineffective. Although orally administered ethanol
has some sleepinducing properties, it often promotes sleep disturbance as the night progresses.
MECHANISM OF ACTION
GABAA Agonists: Benzodiazepines and the Newer Nonbenzodiazepines
When the benzodiazepines first gained prominence in the 1960s, most theories of how
they might act were inherited from previous studies of ethanol and barbiturates, and involved
possible alterations in membrane phospholipid integrity and energy metabolism. With the
discovery of the benzodiazepine receptor (now generally referred to as the GABAA–
benzodiazepine receptor complex) in the late 1970s, attention was focused on the possibility that
the pharmacologic effects of these compounds result from their saturable, stereospecific binding
to a specific recognition site. The receptor complex is part of a group of ligand-gated ion channel
complexes that also includes the glycine, serotonin-3, neuronal nicotinic acetylcholine, and other
receptors. It is functionally comprised of three moieties, a benzodiazepine recognition site, a
GABAA recognition site, and a chloride ionophore.
aDerived from Smith CM, Reynard AM. Essentials of pharmacology. Philadelphia: Saunders; 1995. p. 228, and other sources.
bCitations for kinetic information of benzodiazepines are found in Maczaj .
cOriginally formulated as a hard capsule in the United States, concerns with kinetics and efficacy led to reformulation of the
preparation to a soft gelatin capsule with comparable characteristics of other marketed benzodiazepines of its class.
dNot available in the United States.
eActive metabolite has half-life of 2-5 hrs.
fDrugs that do not have U.S. Food and Drug Administration (FDA) indications for aiding sleep.
gBecause there is not an FDA indication for sleep, there is no FDA recommended dose for this purpose. Doses stated here are
approximations of those often used in clinical practice.
hm-CPP, m-chlorophenylpiperazine.
 Molecular cloning studies indicate that structurally it is comprised of at least five
different glycoprotein subunits (Fig. 42-1). The actual binding site for benzodiazepines appears
to be located at the junction of the alpha and gamma subunits. Each of the subunits may have
multiple isoforms. There appear to be as many as six alpha, three beta, and three gamma
isoforms. Studies using “knock-in” technology in mice indicate that the alpha-1 subunit may
mediate both sedation and memory effects of agonists. This suggests that it will be unlikely to
develop a hypnotic that acts at this site and that will have sleep-promoting, but not amnesic,
properties. Two particular combinations of isoforms that are often noted are the Type-I (alpha 1,
beta 2, gamma 2) and Type-II (alpha 3, beta 2, gamma 2) configurations. The former, the most
common type representing roughly 40% of GABAA receptors, is found in most areas of the brain
and localized in interneurons of the hippocampus and cortex. The latter, perhaps half as
common, is found dominantly in spinal cord motoneurons and hippocampal pyramidal neurons.
Most traditional benzodiazepine hypnotics bind to both types. Some of the newer
nonbenzodiazepine agents, including zolpidem and zaleplon, bind with relatively greater
specificity to the Type-I receptor. Whether this selective binding translates into more specific
pharmacologic properties continues to be studied.

Figure 42-1 Schematic representation of the GABAA–benzodiazepine receptor complex, and possible sites of action of
sleep-inducing drugs. (From Rudolph U, Mohler H: Analysis of GABA-A receptor function and dissection of the pharmacology of
benzodiazepines and general anesthetics through mouse genetics. Annu Rev Pharmacol Toxicol 2004;44:475-498, with
permission.)

The end result of a benzodiazepine agonist binding to its specific site is that as an
outcome of a complex interaction with the GABA recognition site, the flow of negatively
charged chloride ions into the neuron is enhanced, which changes the postsynaptic membrane
potential and alters input resistance such that the postsynaptic neuron is less likely to achieve an
action potential. This inhibitory mechanism is a very potent one in the central nervous system
(CNS), because the GABAA receptor is the most widespread receptor mechanism in inhibitory
synapses, comprising up to 30 percent of all synapses in the CNS. One of the intriguing features
of this view is that it provides a beginning for understanding a long-standing puzzle—how
hypnotic medications of many different pharmacologicclasses may have relatively similar effects
on the process of inducing sleep.
As mentioned previously, the newer nonbenzodiazepine agents such as zolpidem and
zaleplon bind to a subclass of benzodiazepine recognition sites, ethanol has profound effects on
chloride channel function,and barbiturates bind to a distinct site. Barbiturates, for instance, may
cause chloride channels to open for prolonged periods, whereas benzodiazepines may increase
the frequency of opening. Similarly, the active metabolite of chloral hydrate and the anesthetic
propofol modulate GABAA-receptor function. Several lines of evidence have also suggested that
the hypnotic effects of benzodiazepines may involve presynaptic effects mediated by alterations
in potential-dependent calcium channel activity. The calcium channel blocker nifedipine, for
instance, can block the sleep-inducing effects of microinjections of triazolam into the medial
preoptic area. Currently, at least two hypnotics under development address the function of
auxiliary alpha-2 delta subunits of voltage-gated calcium ion channels (Table 42-2).
The original characterization of the receptor complex indicated that it mediates the
anxiolytic, muscle relaxant and anticonvulsant effects of benzodiazepines. That it also mediates
the sleep-inducing effects was demonstrated in a series of studies showing, for instance, that
some beta-carboline compounds that act as receptor inverse agonists induce increased
wakefulness, and at low doses they block sleep induction by benzodiazepines. The binding of
benzodiazepines is stereospecific, such that an enantiomeric form (the B10 compounds) have
opposite effects, one induces sleep whereas the other promotes wakefulness.
Neuroanatomic Considerations
In contrast to the growing understanding of the interaction of benzodiazepine agonists at
a molecular level, the anatomic sites at which they act to induce sleep has been more poorly
understood. The most parsimonious view would be that hypnotics act at loci thought to be
important in sleep regulation on the basis of lesion or stimulation studies (a general review of the
neuroanatomy of sleep is found in Chapter 7). Using that approach, Mendelson microinjected
triazolam into a number of such sites. One of the most striking findings was how in many areas
injections produced no effect on sleep (e.g., the locus coeruleus, the gigantocellular tegmental
fields, the basomedial nucleus of the amygdala, and the ventrolateral preoptic area), or actually
enhanced wakefulness (the dorsal raphe nuclei). In contrast, injections into the medial preoptic
area (MPA) of the anterior hypothalamus consistently enhanced sleep, an anatomically very
specific effect insofar as injections into nearby structures (the lateral preoptic area, the horizontal
limb of the diagonal band of Broca) had no effect. The basal forebrain and anterior hypothalamus
have long been thought to have an important role in sleep regulation.
The MPA is a complex structure that receives afferents from many areas in the forebrain
and brainstem. Among these are projections from various areas of the hypothalamus, as well as
serotonergic fibers from the dorsal raphe nuclei and noradrenergic projections. Cell bodies and
fibers of different parts cross react in immunohistochemical studies with a number of
neuromodulators such as Substance P and Neuropeptide Y. Push-pull cannula studies have
reported release of catecholamines, GABA, and glutamate. GABA is uniformly distributed
throughout the hypothalamus, and its synthetic enzyme is found in high concentrations in the
preoptic area. GABAA–benzodiazepine receptors appear in significant concentrations here,
suggesting that benzodiazepines might inhibit neuronal activity. The preoptic area also contains
cells that are responsive to temperature, osmolarity, glucose, and steroids, and it receives
afferents from various sensory systems.
The basal forebrain contains neurons that increase firing during NREM (non–rapid eye
movement) sleep compared to waking (“sleep-active neurons”), although higher concentrations
are found in the lateral preoptic area and diagonal band of Broca’s area. There are also neurons
that selectively increase or decrease firing rates during REM (rapid eye movement) sleep and
NREM sleep. Thus the preoptic area in general, and the MPA in particular, appear to have a role
in coordinating various systems involved in reproductive and homeostatic functions. Given the
bidirectional interactions of sleep with cardiovascular, thermoregulatory, endocrine, and sensory
systems, it seems likely that the preoptic area might be involved in coordinating sleep with other
systems. It may be, then, that the preoptic area is among the loci at which benzodiazepine
agonists act to induce sleep. Recent work has also indicated that microinjections of pentobarbital
and other GABA receptor agonists into a brainstem area, the mesopontine tegmental anesthesia
area, induced an anesthesia- like state, raising the possibility that it may be involved in drug-
induced loss of consciousness.

Other Neurotransmitters
Subsequent studies have also indicated that microinjections of pentobarbital,29
ethanol,30 adenosine,31 and propofol 32 into the MPA enhance sleep. Indeed, looking at a
somewhat broader area, adenosine has been shown to accumulate extracellularly in the basal
forebrain cholinergic region,33 which in turn has efferents to cortical and thalamic systems
involved in arousal; it has been suggested that alteration of functioning of this cholinergic area
may be the mechanism by which the propensity to sleep is enhanced by prolonged wakefulness.
Similarly, it may be that possible some of the sleep-promoting effects of antihistamines and
tricyclic antidepressants (many of which have significant anticholinergic properties) might be
mediated by alterations of cholinergic neurons in the basal forebrain. The histaminergic system,
which is centered in the tuberomammillary nucleus of the lateral hypothalamus, sends efferents
to the preoptic area as well as the perifornicular area and the cortex, so it also seems possible that
sedative effects of antihistamines and some tricyclic antidepressants might be mediated by
altering the influence of the tuberomammillary nuclei on these centers. It has also been observed
that microinjection of the GABA receptor antagonist gabazine into the tuberomammillary nuclei
inhibits the effects of centrally administered GABAergic agents, suggesting that this area may be
involved in their pharmacologic actions.
There has been increasing interest in the hypocretin/orexin system, which is centered in
the perifornicular region of the posterior and lateral hypothalamus, as a mechanism of arousal.
There also is evidence that various defects in this system are associated with the genesis of
narcolepsy (see Chapter 84). Microinjection of triazolam into this area shortens sleep latency and
increases total sleep time in rats, raising the possibility that such hypnotics reduce wakefulness,
either by direct action at the perifornicular area or indirectly via GABAergic outputs from the
preoptic area. In summary, the manner in which hypnotics may function to induce sleep is not
fully elucidated, but a number of findings suggest that the basal forebrain and anterior
hypothalamus will be among the crucial areas to consider.
CLINICAL EFFECTS
Moving from the neuroanatomic to the clinical level, less is understood about the
mechanism by which hypnotics act. One approach raises the possibility that, among their actions,
hypnotics alter the perception of sleep and wakefulness. This notion grew out of the classical
observation by Rechtschaffen that poor sleepers, when experimentally awakened early in stage 2
sleep, tend to report that they had been awake, while good sleepers tend to report that they had
been asleep. Later studies by Mendelson, replicated this observation and demonstrated that after
administration of triazolam, flurazepam, or zolpidem insomniacs were more likely to report that
they believed they had been asleep compared to when they were given placebo. In contrast, when
flurazepam or zolpidem were given to normal subjects, this effect was not evident.One
interpretation of these data is that hypnotics such as triazolam and zolpidem may correct a
misperception of sleep in some insomniacs, such that their experience of whether they are awake
or asleep becomes more like that of good sleepers.

Other Sleep-Inducing Agents

Most over-the-counter hypnotics are first-generation antihistamines including
diphenhydramine and doxylamine. Although they possess to varying degrees other properties
including anticholinergic effects, their common quality is inhibition of the histamine-1 receptor.
(Later generations of antihistamines are more hydrophilic, and in principle enter the CNS less
readily, thus producing less sedation.) Tolerance to daytime sleepiness appears to develop
rapidly, in about 4 days. As discussed earlier, antihistamines may produce drowsiness by
inhibiting the histaminergic pathways centered in the tuberomammillary nucleus of the posterior
hypothalamus, which enhance wakefulness. It should also be noted that some tricyclic
antidepressants, notably doxepin, which at the time of this writing is under development as a
possible hypnotic, have significant antihistaminic (H1 and H2) properties (although also having a
number of other effects, including antagonism at alpha1 adrenoreceptors and muscarinic
cholinergic receptors, and binding to serotonin 2a and 2c receptor subtypes).
Although there is debate about the effectiveness and range of side effects of melatonin as
a hypnotic, there is now a potent agonist of melatonin subtypes I and II receptors available
(ramelteon), and others in development (agomelatine, TIK-301, VEC-162). It has been
hypothesized that melatonin and agonists might act to affect circadian systems via effects on
melatonin receptors in the suprachiasmatic nucleus, but the ways in which it might alter sleep
have not been well understood. It has been suggested that binding of agonists to the melatonin
type I receptor decreases the waking signal from the suprachiasmatic nucleus. One possibility is
that sleep-promoting effects of agonists at melatonin receptors are at least in part GABAergic.
Melatonin administration is known to raise GABA concentrations in the rat hypothalamus, as
well as 3H-diazepam binding in the forebrain. Similarly, decreases in motor activity produced by
melatonin in the hamster are prevented by the benzodiazepine receptor blocker flumazenil. The
author’s laboratory has reported that microinjections of melatonin into the medial preoptic area
of the rat hypothalamus enhance sleep, suggesting that its site of action may be similar to that of
benzodiazepines, barbiturates, adenosine, and ethanol, as described previously.
Although the topic of anesthetics is beyond the scope of this chapter, we will briefly
mention the intravenous agent propofol, which has made it much more practicable to induce
anesthesia for prolonged periods in, for instance, intensive care unit settings. Neurochemically,
its mechanism of action has not been fully elucidated, though it is known to interact with the
GABAA–benzodiazepine receptor complex, with resultant decreases in acetylcholine release
from the frontal cortex and hippocampus, and to also increase functional activity of dopamine
and serotonin in the cortex. Interestingly, microinjection of propofol into the medial preoptic
area of rats induces sleep, and this effect is blocked by the benzodiazepine receptor blocker
flumazenil.
PHARMACOKINETICS
Benzodiazepines
With the possible exception of temazepam, the benzodiazepines used as hypnotics are
rapidly and completely absorbed, with most achieving peak plasma levels in 1.0 to 1.5 hours.
Some, notably flurazepam, are detectable primarily only in the form of their active metabolites,
and many display kinetics that reflect enterohepatic circulation. Generally oral administration is
more reliable and complete than intramuscular injection. Although there is significant protein
binding, there are few or no cases in which displacement of other protein-bound drugs is
clinically relevant. Most are very lipophilic, and they rapidly enter the CNS, where
concentrations reflect unbound drug in plasma. Whereas the older, longer-acting agents are
metabolized to active compounds, the shorteracting agents such as triazolam are broken down
into inactive substances (Table 42-1, Video 42-1). The elimination half-lives vary widely, from
the relatively short-acting triazolam, to intermediate agents such as temazepam, to long-acting
substances such as flurazepam (Table 42-2). As will be discussed in Chapter 81, accumulation of
the longer elimination half-life agents when taken nightly has significant bearing on one major
clinical issue—the appearance of daytime residual sedation (Fig. 42-2). Because of the
lipophilicity, many are rapidly redistributed, which may play as important a role in the decline in
CNS effects as metabolism. Another implication of the high lipophilicity is that the volume of
distribution is often increased in the elderly (who tend to have a higher ratio of lipid to muscle),
resulting in an increased half-life. Most are broken down by hepatic microsomal systems and
excreted as conjugated glucuronides. There is no stimulation of the hepatic microsomal systems,
and hence no enhancement of the rate of breakdown of other drugs which undergo the same
metabolic processes. Their own metabolism may be inhibited by some compounds such as
cimetidine and some steroids, and it may be accelerated in people who smoke.

Figure 42-2 A hypnotic with a relatively long elimination half life (e.g., greater than 24 hours) will accumulate during nightly
use, in contrast to an agent with relatively short elimination half-life (e.g., 6 hours). (From Nicholson AN: Hypnotics: clinical
pharmacology and therapeutics. In: Kryger MH, Roth T, Dement WC, editors. Principles and practice of sleep medicine, 2nd ed.
Philadelphia: Saunders; 1989. pp. 355-363, with permission.)
The “Z Drugs”
Zolpidem was the first of the “Z drugs” (zolpidem, zaleplon, zopiclone, eszopiclone),
nonbenzodiazepines that bind to various subtypes of the GABAA receptor. In general, their
clinical effects are similar to those of benzodiazepines, though on the basis of animal studies it
has been argued that they have a wider separation in doses producing hypnotic to nonhypnotic
effects. Several studies have suggested that they have a lower risk of dependence and misuse
than the benzodiazepines. A German prescription event study, for instance, described a rate of
misuse about one third that of benzodiazepines. It has been speculated that this may be due to
less affinity for the alpha-2 subtype, which may be related to abuse potential. In the United States
the Z drugs are considered Class IV Drug Enforcement Administration (DEA) restricted agents,
the same classification as the benzodiazepines.

Zolpidem
Zolpidem, an imidazopyridine compound with relative selectivity for the Type-I
GABAA–benzodiazepine receptor, is rapidly absorbed; due to first-pass metabolism, it has a
bioavailability of 67% after oral administration of doses up to 20 mg. Peak concentrations are
reached after 1.6 hours. Total protein binding is approximately 92%. Absorption is slightly
decreased when taken on a full stomach. It has no pharmacologically active metabolites, and is
eliminated primarily by renal excretion. In the United States it is also marketed as a coated two-
layer tablet, in which the inner layer has a more extended release time (Ambien CR).

Zopiclone and Eszopiclone

Zopiclone, which is on the market in Europe and Asia but not the United States, is a
cyclopyrrolone that acts at the GABAA–benzodiazepine receptor complex, but possibly at a
different binding domain or by producing different conformational changes than the
benzodiazepines. It is rapidly absorbed, with peak plasma concentrations occurring in 0.5 to 2
hours. Bioavailability is about 80%, implying that the first-pass effect is relatively small. It is
very lipophilic and enters rapidly into the central nervous system. Protein binding is
approximately 45%. It has two major metabolites, the N-oxide, which has lower pharmacologic
activity, and the inactive N-desmethyl derivative, which along with various minor metabolites
are excreted primarily by the kidneys and lungs.
Eszopiclone, the S-isomer of racemic zopiclone, is marketed in the United States as
Lunesta. Peak concentrations are achieved in 1 hour, with an elimination half-life of
approximately 6 hours. It is weakly bound to protein, and metabolized via oxidation and
demethylation; its biotransformation is attributed to CYP3A4 and CYP2E1. Eszopiclone has
been found to have effectiveness without tolerance for at least 6 months, and was the first drug in
the United States to have no limitation on duration of administration by the FDA.
Zaleplon
A pyrazolopyrimidine that binds selectively to the benzodiazepine- 1 receptor, zaleplon
is rapidly absorbed after oral administration, with peak concentrations being reached in 1 hour,
and has an elimination half-life of approximately 1 hour. Protein binding is approximately 60%.
It is rapidly metabolized to inactive forms, with approximately 71% of labeled compound
recovered in the urine and 17% in feces. Recommendations for administration in the United
States include taking it immediately before bedtime or after the patient has gone to bed and has
experienced difficulty falling asleep. In the latter case, it should be taken at least 4 hours before
time of arising to avoid any possible memory difficulties.

Ramelteon
Ramelteon is a potent agonist at the melatonin types I and II receptors, with negligible
affinity for the GABAA, dopamine, serotonin, or muscarinic cholinergic receptors. It reaches
peak plasma concentrations in 0.75 to 0.94 minutes, and has an elimination half-life of 1 to 2.6
hours; an active hydroxylated metabolite (M-II) has a slightly longer halflife of 2 to 5 hours. It is
metabolized in the liver, primarily with CYP1A2 and to a lesser degree CYP2C9 and CYP3A4.
Its mechanism of action is not fully understood, but it has been hypothesized that binding at
melatonin receptor subtypes in the suprachiasmatic nucleus attenuates its waking signal.
Clinically its efficacy is primarily in decreasing sleep latency, with little effect on awakenings
during the night. It has essentially no dependence-producing effects, and in the United States, it
is not a DEA-restricted drug.

PHARMACOLOGIC PROPERTIES
Although the hypnotic benzodiazepines are given for purposes of aiding sleep, they share
a spectrum of pharmacologic properties with agents given as daytime sedatives or anxiolytics,
indeed, some authors have suggested that the designation of some benzodiazepines as hypnotics
is as much a marketing plan as it has been a pharmacologic decision. Among their effects are
anxiolytic, myorelaxant, and anticonvulsant properties. Many, particularly the longer- acting
agents, have mild respiratory depressant properties, which are much less evident in shorter-acting
agents. There is even some limited evidence that triazolam may improve sleep-disturbed
respiration in central sleep apnea, although whether this is related to direct respiratory effects or
secondary to decreasing the number of arousals (and hence postarousal respiratory pauses)
during sleep is not clear. Even for the longer-acting agents, however, respiratory depression is
much milder than those of older nonbenzodiazepines such as the barbiturates. In practical terms
there is no significant effect in patients with normal ventilation, although it may become evident
when there is preexisting compromised respiration such as in patients with chronic obstructive
pulmonary disease or persons with unrecognized sleep-disordered breathing. The newer
nonbenzodiazepines appear to have very few respiratory effects.
One study of clinically-used doses of zaleplon, for instance, found few effects on
measures of sleep-disturbed respiration in patients with mild-to-moderate obstructive sleep apnea
on continuous positive air pressure. In most cases, the preceding generalizations in this section
apply as well to zolpidem. There is some evidence from animal studies that it possesses a greater
separation of hypnotic and sedative doses. It has been reported to have no respiratory depressant
properties up to doses of 10 mg and to exhibit very mild inhibition of mean inspiratory drive at
20 mg. Zopiclone in therapeutic doses appears to have no significant effect on sleep-disordered
breathing in patients with chronic obstructive pulmonary disease.
In general, adverse reactions to hypnotics are relatively rare and mild; one review of 3
years’ experience in a 1000- bed teaching hospital found the median rate of reported adverse
events to be 0.01% (1 in 10,000) doses administered and ran as high as 0.05%. The rate for
triazolam was 0.02%.
Unlike older hypnotics such as the barbiturates, the hypnotic benzodiazepines are
relatively benign in overdose when taken alone by a medically healthy individual. They may be
very toxic, however, when taken in combination with other central nervous system depressants
such as alcohol, and because a significant portion of overdoses involve a combination of drugs, it
is wise to treat them as potentially toxic or even lethal agents. In practice this translates into
being very conscious of the possibility that a patient seeking help for sleep disturbance may be
suffering from unrecognized depressive illness, and if it is present, initiating appropriate
antidepressant therapy.
EFFECTS ON SLEEP
Polygraphic studies of benzodiazepines indicate that, consistent with their clinical effects,
sleep latency and wake time after sleep onset are generally reduced, and total sleep time is
increased. As with barbiturates and ethanol, spindle activity may be increased. REM sleep time
may be reduced mildly, in contrast to the very potent REM suppression induced by barbiturates.
In the early years after the introduction of benzodiazepines into the U.S. market, much was made
of this observation, which was interpreted to mean that they somehow produced a more natural
sleep. With hindsight, many investigators recognize that the psychological effects of REM
deprivation are much less clear than originally thought, and that in the case of depressed patients,
REM deprivation may even be therapeutic (see Chapter 8), so whether having only mild REM
suppressant properties translates into some clinical advantage seems uncertain. The
benzodiazepines are, however, in contrast to the barbiturates, potent suppressors of slow-wave
sleep. The same dilemma that arises in terms of REM sleep appears once again: Because the
function of slow-wave sleep has not been clearly determined, the clinical significance of
pharmacologically suppressing this stage remains uncertain.
The nonbenzodiazepine zolpidem shares with the benzodiazepines the very mild effects
on REM sleep, but in contrast does not alter slow-wave sleep, which may even increase toward
more expected values in some patients with insomnia with low baseline levels. The development
of neuroimaging techniques is also beginning to give insights into how the newer
nonbenzodiazepines may affect the brain. One recent positron emission tomography (PET) study
indicates that following eszopiclone administration, there is a more rapid decline in metabolic
activity in the midbrain and pontine reticular formation during the transition from waking to
NREM sleep. One interpretation would be that the drug is reducing the hyperarousal often seen
in insomnia.
Clinical efficacy studies indicate that virtually all these agents potently improve
polygraphic measures of sleep and result in better subjective ratings of sleep quality during short-
term use. One of the few distinctions among the benzodiazepines is that the longer-acting agents
such as flurazepam may not have as much effectiveness on sleep latency until the second night of
administration. One concern that was initially raised in terms of the shortacting agents such as
triazolam was the possibility that the relatively rapid metabolism might lead to sleep disturbance
after several hours; later studies analyzing awakenings in the latter part of the night have
generally found no evidence that this is the case.
Another pharmacologic property to consider is the potential for dependence. A review of
this topic by a panel from academia, industry, and the government concluded that the
dependence potential of currently available hypnotics in patients without a history of substance
abuse is minimal. Ramelteon has been found to have no dependence- producing properties in
standard measures. An excellent review of abuse liability and toxicity of 19 hypnotics from the
days of the barbiturates through ramelteon can be found in Griffiths and Johnson. A variety of
other issues at a clinical level include determination of whether or not tolerance develops during
long-term nightly administration, and the effectiveness of nonnightly use, which are considered
in Chapter 81.

FUTURE HYPNOTICS
This is a very fruitful time in the development of hypnotics. Among compounds under
development are ones that act as antagonists to specific serotonin receptor subtypes, alpha-2-
delta calcium channel blockers and orexin antagonists (Table 42-2, Video 42-1). The serotonin
2-subtype receptor has been of interest for some time in relation to the action of some
antidepressants and most atypical antipsychotics; serotonin subtype 2A antagonists including
eplivanserin, pimavanserin and others are under development as hypnotics, with studies
emphasizing improvements in sleep continuity and minimal daytime sedation. There are also
compounds which combine antagonism at serotonin 2C receptors with melatonin receptor
agonism.
Compounds that bind to alpha-2-delta subunits of voltage-gated calcium channels, such
as the anticonvulsants pregabalin and gabapentin, increase calcium ion entry into neurons and
alter GABA concentrations. Both appear to increase slow-wave sleep and decrease awakenings,
and are under development as hypnotics. Orexin/hypocretins are excitatory hypothalamic
neuropeptides that enhance waking, and indeed a deficit in orexin-containing neurons is related
to the genesis of narcolepsy (see Chapter 84), an antagonist to orexin receptors, GW649868, is
under development as a hypnotic as well. As mentioned earlier, low dose administration of the
tricyclic antidepressant doxepin, which has significant antihistaminergic properties as well as
antagonism at alpha1 adrenoceptors and other properties, is being developed to aid sleep.
Advances in medicinal chemistry are enabling the creation of hypnotics that may come in
multiple preparations varying in duration of action. There has been interest in novel modes of
administration including inhalation, oral sprays and oral dissolving tablets.