Pharmacologic Agents for Insomnia

Pharmacologic Treatments

A variety of pharmacologic treatments are available for the management of insomnia. Categories of prescription medications commonly used for this purpose are listed in Table 11-1. Various pharmacokinetic properties, including rate of absorption, extent of distribution and elimination half-life, determine important clinical effects, such as onset and duration of action and the predilection for daytime carryover effects. Optimally, the hypnotic agent selected should be tailored to the nature of the patient’s specific insomnia complaint and existing comorbidities, if any.

Sleep variables are summarized in Figure 11-1. Sleep latency is a measure of the time spent falling asleep after going to bed; wake after sleep onset (WASO), the total time spent awake after falling asleep, is a measure of sleep discontinuity during the course of the night; and total sleep time (TST) is a measure of the duration of sleep.

For patients having difficulty only falling asleep, the ideal medication would reach peak plasma concentrations rapidly, thus producing a rapid onset of action and then be rapidly eliminated. For patients who have difficulty with sleep maintenance, i.e., multiple nocturnal awakenings or early morning awakening, an ideal medication would allow for mid-nocturnal administration, following middle-of-the-night (MOTN) awakening and have a rapid onset of action, as well as a rapid elimination rate to prevent carry-over effects following morning awakening. Another option is a medication that would be taken just prior to bedtime, and have a sufficiently long elimination half-life to ensure sleep maintenance during the problem period, yet have no residual effects following morning awakening. Finally, for patients having difficulties with both sleep initiation and sleep maintenance, the ideal medication would provide rapid onset of action to help the patient fall asleep and a sufficiently lengthy duration of action to prevent MOTN awakenings, yet have no residual effects.

Historical Aspects

Although a variety of substances have been used for the treatment of insomnia since antiquity, bromide, introduced in the mid 1800s, was the first agent specifically used as a sedative hypnotic. Chloral hydrate, paraldehyde, urethan and sulfonyl soon followed.

Chloral hydrate’s chemical structure is distinct from that of barbiturates and benzodiazepines. Prescribed in doses of 500 to 1,000 mg, it has a mean half-life of 4 to 8 hours. It is an effective hypnotic and has a rapid onset of action. Side effects include:

• Gastric irritation
• Unpleasant taste and odor
• Daytime hangover
• Light-headedness
• Malaise
• Ataxia
• Nightmares
• Rarely, paradoxical excitement.

Tolerance develops rapidly, usually within 5 weeks of use. Continued use, especially in large doses, can result in:

• Hypotension
• Arrhythmias
• Myocardial depression
• Physical and psychological dependence
• Hepatic damage.

The lethal to therapeutic ratio is quite narrow, and overdose can result in severe respiratory depression, hepatic damage, hypotension, coma and death. Rapid discontinuation following chronic use is associated with a withdrawal syndrome characterized by delirium, seizures and death. For these many reasons, chloral hydrate is no longer used for insomnia.

The barbiturates (e.g., phenobarbital, secobarbital, butabarbital, amobarbital, and pentobarbital) were introduced in the early 1900s and by the 1960s, accounted for 55% of all hypnotic prescriptions. They produce generalized depression of the CNS, ranging from mild sedation to general anesthesia. Despite their past use as sedative-hypnotic drugs, they have been replaced largely by much safer agents for the treatment of insomnia. Hypnotic doses of barbiturates increase TST, decrease sleep latency, the number of awakenings and the duration of REM and N3 sleep. Tolerance is seen following a few days of administration, and TST may be reduced by as much as 50% after 2 weeks of use. Rapid discontinuation typically leads to rebound effects, including a rebound of REM sleep, associated with increased frequency and intensity of dreaming and nightmares. Barbiturates are associated with residual sedation (hangover) and psychological dependence. Barbiturate overdose, especially when it also involves alcohol, can be fatal. For these many reasons, barbiturates are not commonly utilized for the treatment of insomnia.

FDA-Approved Hypnotics

Benzodiazepine Receptor Agonists

Owing to their relatively low risk of fatal CNS depression, the BzRAs have displaced chloral hydrate and the barbiturates over 3 decades ago as sedative-hypnotic agents. All agents in this class bind to benzodiazepine recognition site of the γ-aminobutyric acid type A (GABA-A) receptor complex and augment the effects of GABA. GABA is the most abundant inhibitory neurotransmitter in the CNS and is thought to mediate a wide variety of clinical effects, including anxiety, cognition, vigilance, memory, balance and learning, among others. GABA is also the major neurotransmitter of neurons in brain structures thought to be critical for the generation of sleep, such as the ventrolateral preoptic nucleus of the hypothalamus.

GABA-A receptors are widely distributed in the CNS, including in the cortex, basal ganglia and cerebellum. These receptors contain not only the GABA receptor itself but also a benzodiazepine recognition site and a chloride ion channel. GABA-A receptors are composed of 5 subunits (α, β, γ, ε and ρ, Figure 11-2). Most GABA-A receptors are composed of 2 α, 2 β, and 1 γ subunits. The BzRAs bind to the benzodiazepine recognition site, located at the interface of α and γ subunits. Each of these subunits exists in multiple forms and different combinations of these forms may yield different pharmacologic properties, although this connection has not been firmly established.

Benzodiazepines act with comparable affinity at all GABA-A receptors containing a β, a γ-2, and any of 4 α subunits (α 1, α 2, α 3, or α 5). They do not interact with receptor subtypes containing an α 4 or α 6 subunit. The newer hypnotic agents, such as zaleplon, eszopiclone and zolpidem, bind more avidly to benzodiazepine receptors containing the α 1 subunit. These agents are collectively referred to as selective benzodiazepine receptor agonists (sBzRAs). In animal models, GABA-A receptors with α 1 subunits are thought to mediate the sedative, amnestic and anticonvulsant effects, whereas those containing α 2 subunits are more important for anxiolytic effects. However, the connection between subunit composition and physiologic activity in humans is still under evaluation.

The BzRAs that are most commonly used are listed in Table 11-2 and Table 11-3. The first group represents agents that have a benzodiazepine molecular structure. The second group represents agents that are nonbenzodiazepines in structure and are the sBzRAs.

As is evident in Figure 11-3 and Figure 11-4, plasma concentrations of available BzRA hypnotics follow a skewed pattern. In general, their plasma concentrations peak rapidly; therefore, with the possible exception of temazepam, all are effective in reducing sleep latency and are suitable for patients who complain of difficulty in falling asleep upon retiring (Figure 11-5). As a group, the older BzRAs are also effective in increasing TST and in maintaining sleep (i.e., reducing WASO), although these effects have not been well demonstrated with all of these agents. Therefore, they are also well suited for patients who complain that they awaken repeatedly during the course of the night or who cannot sleep for adequate periods of time. However, largely owing to their longer elimination half-lives, which result in plasma concentrations that persist into waking periods, the older BzRAs also have a tendency to produce daytime carryover effects.

Zolpidem ER, an extended-release formulation of zolpidem, consists of a coated two-layer tablet: one layer releases drug content immediately and another allows a slower release of additional drug content, thus providing extended plasma concentrations beyond 3 hours after administration (Figure 11-6).

Following the development of the first sBzRA, oral zolpidem, many newer formulations have been subsequently developed. These feature drug delivery via the oral mucosa via administration in the sublingual route, an area of the oral cavity that is more permeable than the cheek and palatal areas. Sublingual administration is useful when rapid onset of action is desired. The portion of drug absorbed through the sublingual blood vessels bypasses the hepatic first-pass metabolic processes, and is protected from degradation by the low pH environment and digestive enzymes of the middle gastrointestinal tract. Similarly, sublingual sprays improve the time to reach maximum plasma concentration compared with other types of sublingual dosage forms.

A sublingual formulation of zolpidem has been developed for the short-term treatment of insomnia characterized by difficulties with sleep initiation. Possibly owing to its rapid absorption, zolpidem SL appears to provide an earlier onset of action than oral zolpidem.

A lower-dose sublingual formulation of zolpidem was formulated specifically for the treatment of patients with insomnia characterized by difficulty returning to sleep following MOTN awakenings. This formulation has a rapid onset of action and short duration of action, which is suitable for patients who wake up in the MOTN with at least 4 hours of bedtime remaining.

The efficacy and safety of low-dose zolpidem sublingual tablets were established in two studies in patients with primary insomnia with a history of prolonged MOTN awakenings (at least three per week of at least 30 minutes in duration). In a laboratory study, each subject completed three treatment periods consisting of two consecutive nights followed by a waiting interval. During each period, patients were awakened 4 hours after bedtime, given zolpidem sublingual tablets or placebo and kept awake for 30 minutes before they were allowed to return to sleep for another 4 hours. Compared with placebo, zolpidem sublingual tablets significantly decreased both objective (polysomnographic) and subjective (patient-estimated) measures of sleep latency. The effect on sleep latency was similar for females receiving 1.75 mg of zolpidem sublingual tablets and men receiving 3.5 mg of zolpidem sublingual tablets. Thus the time for patients to fall back asleep after MOTN awakening was decreased with zolpidem treatment (Figure 11-7), and total sleep time was increased. There was no evidence of daytime impairment.

Additionally, zolpidem sublingual tablets were evaluated in patients with insomnia who experienced difficulty returning to sleep after MOTN awakening. During a 4-week outpatient study, zolpidem sublingual tablets or placebo was taken on an as-needed basis when subjects had difficulty returning to sleep after spontaneous MOTN awakenings. Subjective time to fall back asleep after a MOTN awakening was significantly shorter for zolpidem sublingual tablets 3.5 mg (Figure 11-8). There was no evidence of residual effects on the morning after dosing based on a 9-point scale assessing morning sleepiness/alertness. The zolpidem sublingual tablets were generally well tolerated, and adverse events were generally mild and similar for both groups.

All of the newer sBzRAs also increase TST; however, in the case of zaleplon, the latter feature is limited to the highest dose of 20 mg. With respect to effects on sleep discontinuity, eszopiclone and zolpidem ER also diminish WASO and, therefore, enhance sleep continuity throughout the course of the night (Table 11-4). Eszopiclone was the first hypnotic that was shown to have this effect for as long as 6 months of continuous nightly treatment in a controlled trial (Figure 11-9). This feature may be due to its longer elimination half-life (6 hours) when compared with the other sBzRAs. Zolpidem ER also enhances sleep continuity (decreases WASO), an effect that has been demonstrated in controlled trials. These effects have been noted for the first 7 hours following administration during the first 2 nights and for the first 5 hours following administration after 2 weeks of treatment. This clearly distinguishes it from zolpidem. In one study, patients who were administered zolpidem ER at bedtime compared with those who were administered zolpidem were able to fall asleep more rapidly following experimentally-induced mid-nocturnal awakenings (Figure 11-10).

Most common adverse effects of BzRA hypnotic medications are somnolence, headache, nausea, fatigue, hypokinesia, dizziness and abnormal coordination. Severe sedation, nervousness, lethargy, dry mouth, diarrhea and coma, probably indicative of drug intolerance or overdosage, have been reported. Similarly, the most commonly reported adverse events of sBzRAs include headache, nausea, fatigue, drowsiness, dizziness, diarrhea and drugged feeling. These occur variably in each of the medications (refer to individual product inserts for detailed information on each medication).

All of the BzRAs are classified as Schedule IV agents by federal regulation (the US Drug Enforcement Agency) and carry the risk of abuse liability (see Dependence and Abuse discussion, below). They should, therefore, be used with special caution in individuals with a prior history of alcohol and substance abuse. Adequate and well-controlled studies do not exist regarding pregnancy or lactation in humans for the sBzRAs; therefore, their benefits for usage in pregnancy and in lactating women must be weighed against their risks. None of these medications are indicated for pediatric usage. Prescribing guidelines for the BzRAs describe the risk of complex behaviors such as “sleep-driving” (i.e., driving while not fully awake after ingestion of a sedative-hypnotic, with amnesia for the event), which increase in likelihood with the use of alcohol and other CNS depressants as well as doses exceeding the maximum recommended dose.

Melatonin Receptor Agonists

As noted in NORMAL SLEEPthe sleep-wake cycle represents one of many circadian rhythms that are regulated by an internal biologic clock, which is located in the suprachiasmatic nucleus (SCN) of the anterior hypothalamus. During the course of a typical day, sleep-producing homeostatic factors (sleep debt or sleep pressure) accumulate and intensify in force as the day progresses. In order to maintain the waking state, the SCN produces an opposing wakefulness force, which also intensifies as the day progresses, and which counteracts the homeostatic factor, thus resulting in the maintenance of wakefulness and normal daytime functioning. Localized to the SCN are melatonin receptors, sites at which endogenous melatonin binds and affects the neuronal output of the SCN.

There are two subtypes of such receptors, MT₁ and MT₂, and these are G–protein-coupled receptors. Activation of MT₁ receptors inhibits the neuronal firing rate in the SCN, thus tipping the scale in the direction of the homeostatic sleep-promoting factors. MT₂ receptor activation may play a role in readjustment of circadian rhythms. Throughout the 24-hour cycle, receptor sensitivity fluctuation can be modified by either endogenous melatonin or by heterologous factors. Melatonin receptor agonists act at the MT₁ and MT₂ receptors, and, possibly owing to their activity at the MT1 receptors, are thought to mute the wakefulness force of the SCN, thus allowing the homeostatic factor to dominate, resulting in sleep.

Ramelteon is the only melatonin receptor agonist available for clinical use and specifically targets MT₁ and MT₂ receptors with high selectivity, and has virtually no binding capacity to MT₃ receptors located in numerous locations outside the CNS.

Ramelteon is indicated for the treatment of insomnia characterized by difficulty with sleep onset (Table 11-5).

Ramelteon has been shown to reduce patient-reported sleep latency in 829 older adults with chronic insomnia. The improvements were sustained throughout the 5-week, double-blind treatment period (Figure 11-11), and no significant rebound insomnia or withdrawal effects were observed. In a subset analysis of older adults with primary chronic insomnia characterized by severe baseline sleep-onset difficulties (subjective sleep latency ≥60 minutes) who had participated in the above trial, ramelteon 8 mg persistently reduced subjective reports of time to sleep onset during 5 weeks of nightly treatment. Ramelteon 8 mg was also shown to reduce sleep latency and increase TST in a placebo-controlled, sleep laboratory study in 289 adults with transient insomnia. Ramelteon did not affect WASO or number of awakenings. No significant next-morning residual effects were observed.

The most common adverse events that are associated with ramelteon include somnolence, fatigue and dizziness. It is not recommended for use with fluvoxamine due to a CYP 1A2 interaction. A mild transitory elevation in prolactin levels has been noted in a small number of women and a mild decrease in testosterone values has been noted in elderly men, yet the clinical relevance of these changes remains unclear. Possibly owing to its lack of activity at the GABA receptor, ramelteon does not demonstrate respiratory depression in mild to moderate OSAS or in mild to moderate COPD.

Data for ramelteon use in pregnant and lactating women are lacking. It is not scheduled by the DEA as a controlled substance. Its usage is not indicated in pediatric population.

H1-Receptor Antagonists

Low-dosage (3 mg and 6 mg) formulations of the tricyclic antidepressant doxepin are FDA-approved for the treatment of insomnia characterized by frequent or early morning awakenings and an inability to return to sleep (Table 11-5). Doxepin, like ramelteon, is not designated as a controlled substance and thus may be of special value in patients with a history of substance abuse.

Although the specific mechanism of action of doxepin for treating insomnia remains unknown, its blockade of the H₁ receptor likely plays a role in reducing wakefulness. Compared with other tricyclic antidepressants, the binding potency to the H₁ receptor of doxepin is approximately 100 times higher than its binding potency for serotonin and norepinephrine. Thus at low doses, doxepin acts as a highly selective H₁ antagonist.

The efficacy of doxepin 1 mg, 3 mg and 6 mg for reducing insomnia symptoms was assessed in two phase 3 randomized, double-blind, placebo-controlled clinical trials using a 4-period crossover design. One trial enrolled 67 patients, ages 18 to 64 with chronic primary insomnia, while the second trial included 76 adults age ≥65 years with primary insomnia. In both trials, patients were randomly assigned to placebo or 1 mg, 3 mg, or 6 mg of doxepin for two nights. All patients received all treatments. Each treatment was followed by 8 hours of PSG evaluation in a sleep laboratory.

Wake time during sleep (WTDS) was the primary efficacy endpoint is both trials. In the first study, patients taking doxepin at all doses achieved improvement in objective (PSG-defined) and subjective (patient-reported) measures of sleep duration and sleep maintenance. WASO, TST and sleep efficiency improved with all doxepin doses, while WTDS increased with the 3-mg and 6-mg doses, but not with 1 mg or placebo. In addition, PSG indicators of early-morning awakenings (terminal insomnia) were reduced. Improved sleep onset was seen only with the 6-mg dose. In the second trial, patients taking any doxepin dose also achieved objective and subjective improvement in sleep duration and sleep maintenance, which lasted into the final hours of the night. In both studies, next-day alertness was assessed using the VAS for sleepiness, and the Digit-Symbol Substitution Test and the Symbol-Copying Task for psychomotor function. In both studies, there were no statistically significant differences in these measures between treatment with any of the doxepin doses or placebo.

Doxepin appears to be better tolerated at hypnotic doses (3 mg and 6 mg) than at antidepressant doses (50 to 300 mg/d), although direct comparative studies are not available. In the two phase 3 clinical studies that lasted up to 3 months, the overall safety profiles of all three doses of doxepin were comparable to placebo. There were no differences in next-day sleepiness or psychomotor function among patients who received placebo or doxepin.

Clinical studies that evaluated the safety of doxepin lasted up to 3 months. Somnolence/sedation, nausea and upper respiratory tract infection were reported by >2% of patients taking doxepin and were more common than in patients treated with placebo.

There are no adequate and well-controlled studies of doxepin in pregnant women. Doxepin is excreted in human milk after oral administration. It is not scheduled as a controlled substance and is not indicated for pediatric use. Doxepin is contraindicated in patients with hypersensitivity to doxepin hydrochloride, severe urinary retention, narrow angle glaucoma, and in those who have used monoamine oxidase inhibitors (MAOIs) within the previous 2 weeks.

Orexin Receptor Antagonists

As noted in NORMAL SLEEP, the neuropeptide hypocretin/orexin plays an important role as a stabilizer and maintainer of wakefulness, minimizing unplanned transitions to the sleep state through the reinforcement of wake-promoting signaling in the brain. Orexin deficiency results in narcolepsy in many species, suggesting that this system is particularly important for maintenance of wakefulness, although not necessarily its initiation. Orexin peptides bind with different affinities to the two orexin receptors, OX1R and OX2R, to induce downstream wake signaling. Several compounds that function as antagonists at one or both of these receptors have been identified as possible insomnia treatments, based on the notion that reducing orexin activity might benefit excessively wakeful insomnia patients by allowing them to achieve sleep.

Suvorexant is the first orexin antagonist to be approved for the treatment of insomnia. By antagonizing orexin receptors, it is presumed to dampen downstream wake-promoting signaling and in turn allowing for sleep-promoting mechanisms to take effect. Suvorexant is indicated for the treatment of insomnia characterized by difficulties with sleep onset and/or sleep maintenance. Suvorexant peak concentrations occur at a median T_max of 2 hours (range 30 minutes to 6 hours) under fasted conditions. The mean absolute bioavailability of 10 mg is 82%. The primary elimination route is fecal, 66% and 23% is eliminated through the urine. No dosage adjustments are needed for those with renal impairment. Steady-state is achieved 3 days after ingestion. The mean t_1/2 is approximately 12 hours.

The efficacy and safety of suvorexant has been established at doses of 10 mg to 20 mg in elderly and non-elderly adult patients. The recommended starting dose is 10 mg for most patients, taken no more than once per night and within 30 minutes of going to bed, with at least 7 hours remaining before the planned time of awakening. However, 15-mg and 20 -mg doses could be appropriate in patients in whom the 10-mg dose is well-tolerated but not effective. For patients taking concomitant moderate CYP3A4 inhibitors, a 5-mg dose may be necessary.

In two 3-month controlled efficacy trials, Study 1 and Study 2, patients with insomnia characterized by difficulties with sleep onset and sleep maintenance were exposed to suvorexant 15 mg or 20 mg or placebo. Suvorexant 15 mg and 20 mg were found to be superior to placebo for both sleep latency and sleep maintenance, as assessed objectively by PSG (Figure 11-12 and Figure 11-13) and subjectively by patient-estimated sleep latency and TST (Figures 11-14 and Figure 11-15). These effects did not change over time and were similar between women and men and, based on limited data, between Caucasians and non-Caucasians as well.

In phase 3 clinical studies, the orexin receptor antagonist suvorexant was well-tolerated at 0-3 months, with 22.1% and 15% of patients experiencing drug-related adverse events in the suvorexant and placebo groups, respectively.

In clinical trials of suvorexant, the most common adverse reaction was somnolence, 7% compared with placebo, 3%. The discontinuation rate due to adverse reactions for both 15 mg or 20 mg was 3% compared with 5% for placebo. No individual adverse reaction led to discontinuation at an incidence ≥1%. A small number of sleep-related hallucinations, sleep paralysis and complex sleep-related behaviors were reported by patients taking suvorexant, and although reported in previous trials of other sedative hypnotics, the product insert does caution of this possibility, as well as cataplexy. Suvorexant was associated with a dose-related increase in somnolence, 2% at the 10-mg dose and 5% at the 20-mg dose compared with <1% for placebo.

Suvorexant is a DEA Schedule IV medication and is, therefore, a controlled substance. It has not been well studied in the context of pregnancy and lactation in humans. It is not indicated in pediatric populations. It is contraindicated in patients with narcolepsy.

Lemborexant and daridorexant are the most recently introduced hypnotic agents and readers may be less familiar with them. We will, therefore, provide greater detail regarding these two agents. Lemborexant is an orexin receptor antagonist indicated for the treatment of adult patients with insomnia characterized by difficulties with sleep onset and/or sleep maintenance. It was approved by the FDA in 2019. Lemborexant reaches peak plasma concentrations in approximately 1-3 hours (T_max) when fasting; if taken after a high-fat and high-calorie meal, T_max is delayed by 2 hours. Lemborexant is primarily eliminated through feces (57.4%) and urine (29.1%). Lemborexant 5 mg and 10 mg have an effective t_1/2 of 17 and 19 hours, respectively.

The efficacy and safety of lemborexant 5 mg and 10 mg was evaluated in elderly and non-elderly patients who met DSM-5 criteria for insomnia disorder. The recommended dosage for lemborexant is 5 mg given immediately before bed, no more than once per night and at least 7 hours before the intended time of awakening. Depending on the clinical response and tolerance, the dose may be increased to a maximum of 10 mg. If taken with or right after a meal, the onset of sleep may be delayed. The maximum dose for patients with moderate hepatic impairment is 5 mg, and use in patients with severe hepatic impairment is not recommended.

Approval of lemborexant was based on data from two randomized, double-blind, phase 3 clinical trials: SUNRISE 1, which was placebo- and active comparator-controlled, and recruited a population of older patients and SUNRISE 2, which was placebo-controlled and designed to test the long term safety and efficacy. The SUNRISE 1 trial consisted of a 30-day treatment period and 14-18 days of follow-up; SUNRISE 2 involved a 6-month placebo-controlled treatment period followed by a 6-month active-treatment-only period. In SUNRISE 1, the primary endpoint was the change from baseline in the latency to persistent sleep (LPS) for lemborexant compared to the placebo (assessed at nights 1/2 and 29/30); the key secondary endpoints were changes from baseline in sleep efficiency and WASO vs placebo, and WASO in the second half of the night (WASO2H) vs zolpidem (all assessed at nights 1/2 and 29/30). The primary endpoint of SUNRISE 2 was mean change from baseline in subjective sleep onset latency (sSOL), while mean changes from baseline in subjective sleep efficiency (sSE) and subjective WASO (sWASO) were the key secondary endpoints (all assessed at 6 months).

In SUNRISE 1, a total of 1006 older adult male (≥65 years) and female (≥55 years) patients were randomized (5:5:5:4) to receive lemborexant 5 mg, lemborexant 10 mg, zolpidem 6.25 mg, or placebo. At the end of one month of treatment (nights 29 and 30), both doses of lemborexant were superior to the placebo with respect to changes from baseline in LPS, sleep efficiency, and WASO, and superior to zolpidem with respect to WASO2H (Figure 11-16). In SUNRISE 2, a total of 971 adult (≥18 years) male and female patients were randomized (1:1:1) to receive lemborexant 5 mg, lemborexant 10 mg, or placebo. At 6 months, lemborexant was superior to the placebo at both doses with regards to baseline changes in sSOL, sSE, and sWASO (Figure 11-17).

Comparable treatment-emergent adverse event incidence rates were reported across all treatment groups in SUNRISE 1 (lemborexant 5 mg: 27.8%; lemborexant 10 mg: 30.6%; zolpidem 6.25 mg: 35.4%; placebo: 25.4%) and SUNRISE 2 (lemborexant 5 mg: 61.1%; lemborexant 10 mg: 59.6%; placebo: 62.7%). The most common adverse reactions reported in the first 30 days of SUNRISE 1 and 2 (combined data) in at least 2% of patients and at a higher frequency with lemborexant than with placebo included somnolence or fatigue (lemborexant 5 mg: 6.9%; lemborexant 10 mg: 9.6%; placebo: 1.3%), headache (lemborexant 5 mg: 5.9%; lemborexant 10 mg: 4.5%; placebo: 3.4%), and nightmares or abnormal dreams (lemborexant 5 mg: 0.9%; lemborexant 10 mg: 2.2%; placebo: 0.9%). Adverse events which occurred in less than 2% of patients who received lemborexant and no patients who received the placebo included sleep paralysis (lemborexant 5 mg: 1.3%; lemborexant 10 mg: 1.6%), hypnagogic hallucinations (lemborexant 5 mg: 0.1%; lemborexant 10 mg: 0.7%), and complex sleep behaviors (two events among patients receiving lemborexant 10 mg).

Lemborexant is a DEA Schedule IV controlled substance. There is no data on the effects of lemborexant use in pregnant women. Limited data in lactating women suggest a low transfer of lemborexant into breast milk (relative infant dose <2% of the maternal dose). Infants exposed to lemborexant through breastmilk should be monitored for excessive sedation. Lemborexant is contraindicated in patients with narcolepsy.

In 2022, the FDA approved a third orexin receptor antagonist, daridorexant, for the treatment of adult patients with insomnia characterized by difficulties with sleep onset and/or sleep maintenance. Daridorexant reaches peak plasma concentrations within 1–2 hours (T_max) and the its terminal t_1/2 is approximately 8 hours. The absolute bioavailability of daridorexant is 62%. The primary route of daridorexant excretion is via feces, approximately 57%, and about 28% is excreted by urine.

The efficacy and safety of daridorexant at the 10 mg, 25 mg and 50 mg dose was evaluated in adults ≥18 years of age with insomnia disorder (according to the DSM-5). The recommended dosage is 25 mg or 50 mg once per night, taken orally within 30 minutes before going to bed, with at least 7 hours remaining prior to planned awakening. For patients with moderate hepatic impairment, the maximum recommended dosage is 25 mg no more than once per night; daridorexant is not recommended in patients with severe hepatic impairment.

Two randomized, phase 3 clinical trials, Study 1 and Study 2, evaluated the safety and efficacy of daridorexant. These trials consisted of a screening period, a single-blind placebo run-in period, a double-blind treatment period and a 7-day placebo run-out period, after which patients could enter a 9-month, double-blind, placebo-controlled extension study (Study 3). At months 1 and 3, patients underwent polysomnography for the assessment of the primary endpoints of WASO and latency to persistent sleep (LPS). The secondary endpoints of changes from baseline in the self-reported TST and the sleepiness domain score of the Insomnia Daytime Symptoms and Impacts Questionnaire (IDSIQ) were also assessed at months 1 and 3.

During the 3-month-long treatment period, patients received either 50 mg daridorexant, 25 mg daridorexant, or placebo (1:1:1) in Study 1, or 25 mg daridorexant, 10 mg daridorexant, or placebo (1:1:1) in Study 2. WASO and LPS were significantly improved at the end of month 1 and month 3 with both 50 mg and 25 mg dose compared to placebo in Study 1 (Figure 11-18), as was TST (Figure 11-19) while IDSIQ sleepiness domain scores improved only in the 50 mg group (Figure 11-19). In Study 2, participants in the daridorexant 25 mg group showed a significant reduction in WASO but no significant change in LPS (Figure 11-18) at months 1 and 3. A significant improvement in TST was observed, but there was no significant change in the IDSIQ sleepiness domain scores (Figure 11-19). By contrast, no significant difference between daridorexant 10 mg and placebo was found for any of the primary or secondary endpoints (Figure 11-18 and Figure 11-19).

The overall incidence of adverse events was comparable between treatment groups in Study 1 (38% in the daridorexant 50 mg group, 38% in the 25 mg group and 34% in the placebo group) and Study 2 (39% in the daridorexant 25 mg group, 38% in the 10 mg group and 33% in the placebo group).

The most common adverse reactions reported in at least 5% of patients and greater than in the placebo-treated group were headache (6% for daridorexant 25 mg, 7% for daridorexant 50 mg and 5% for the placebo) and somnolence or fatigue (6% for daridorexant 25 mg, 5% for daridorexant 10 mg and 4% for the placebo).^{[Quviviq PI]} Rare adverse reactions included sleep paralysis (0.5% and 0.3% of patients receiving daridorexant 25 mg and 50 mg, respectively, compared to no reports for the placebo) and hypnagogic and hypnopompic hallucinations (0.6% of patients receiving daridorexant 25 mg compared to no cases with daridorexant 50 mg or the placebo). Some post-approval voluntary reports from individuals using daridorexant included abnormal or bad dreams and hypersensitive reactions such as rash/urticaria.

Daridorexant is a Schedule IV controlled substance. There are no data on the effects of daridorexant use in pregnant and/or lactating women; however, daridorexant and its metabolites were found in the milk of lactating rats. Infants exposed to daridorexant through breastmilk should be monitored for excessive sedation. Daridorexant is contraindicated in patients with narcolepsy.

Other Clinical Considerations in the Use of Hypnotic Agents

Long-Term Use

The chronic nature of some patients’ insomnia may necessitate longer treatment. The report of an NIH State of the Science Conference expressed concern regarding the mismatch between the potentially lifelong nature of insomnia and the duration of clinical trials.

The main concerns in long-term use are tolerance, a decrement in clinical efficacy following repeated use, and rebound insomnia, which is an escalation of insomnia beyond baseline severity levels following abrupt discontinuation. The latter must be distinguished from a return of symptoms after discontinuation of the medication. Studies of the repeated administration of BzRA hypnotic agents for 2 to 5 weeks suggest that rebound phenomena following withdrawal are more pronounced following the administration of higher doses of BzRA hypnotics and following administration of the benzodiazepine agents that have a short elimination half-life, such as triazolam, than the longer-elimination half-life benzodiazepines and some of the newer sBzRAs. They are less likely following the administration of long-acting drugs because of the gradual decline in their plasma concentration following discontinuation. Mild and transient withdrawal effects, generally lasting 1 day, have been noted with zolpidem.

In a study utilizing intermittent treatment with zolpidem over the course of 3 months, patients were instructed to take the medication a minimum of three and a maximum of five pills per week. Over the course of the study, clinical gains were sustained and there was no evidence of subjective rebound insomnia on nights when the medication was not taken (Figure 11-20). As noted above, a study utilizing nightly eszopiclone 3 mg for 6 months in a placebo-controlled design (Figure 11-8) revealed no evidence of tolerance during the entire 6-month course; sustained improvement was noted in, among other measures, sleep latency, WASO and TST. Monthly ratings of next-day function, alertness and sense of well-being were also improved. There was also no evidence of rebound insomnia after rapid discontinuation in a subset of patients who received treatment for an additional 6 months in an open-label fashion.

The long-term usage of zolpidem ER was evaluated in a 6-month, placebo-controlled study. Patients self-administered zolpidem ER 12.5 mg or placebo for a minimum of 3 nights/week up to a maximum of 7 nights/week. Zolpidem ER resulted in improvement in all four patient global impression items (Figure 11-21). At every time point, zolpidem ER demonstrated sustained efficacy in TST, WASO, and sleep latency. No rebound insomnia was observed during the first 3 nights after discontinuation and there were sustained improvements in morning sleepiness and ability to concentrate with zolpidem ER compared with placebo.

The long-term usage of zaleplon has been studied from a safety standpoint. In an open-label study, older patients (≥65 years of age) self-administered zaleplon nightly from 6 to 12 months. Adverse events were unremarkable. The study confirmed the relative safety and suggested the efficacy of long-term therapy with zaleplon at doses of 5 and 10 mg.

The long-term efficacy of ramelteon was evaluated in a 6-month randomized, placebo-controlled sleep laboratory study in 459 adults with chronic primary insomnia. Over the 6 months of treatment, ramelteon consistently reduced latency to persistent sleep compared with baseline and with placebo (Figure 11-22), with significant decreases observed at week 1 and months 1, 3, 5 and 6. There were no significant next-morning residual effects during treatment and no withdrawal symptoms or rebound insomnia after discontinuation of ramelteon treatment.

In a long-term, 3-month clinical trial, a 3-mg dose of doxepin given nightly was shown to produce significant improvement in sleep latency without evidence of next-day residual sedation.

In a long-term study, the safety and efficacy of suvorexant during 1-year treatment of insomnia was evaluated, along with subsequent abrupt treatment discontinuation, in a randomized, placebo-controlled, parallel-group trial. Patients were all 18 years of age and older, with 59% in both arms being ≥65 years old. 322 receiving the medication completed the 1-year study compared to 162 taking placebo. Dosages of suvorexant were above those approved for clinical use; the dose of suvorexant was 30 mg nightly for elderly patients and 40 mg nightly for nonelderly patients, both higher than the maximum approved dosage of 20 mg. The primary objective for the 1-year phase was to assess the safety and tolerability of suvorexant. Over the 1-year period, suvorexant was considered to be generally safe and well tolerated (Table 11-6).

Currently available long-term safety and efficacy data for lemborexant come from SUNRISE 2, which included exposure to lemborexant for up to 12 months. In patients who received 12 months of lemborexant treatment, the sSOL, sSE and sWASO improvements observed at 6 months were maintained at 12 months (Figure 11-17). No evidence of rebound insomnia or withdrawal was noted following lemborexant discontinuation. The safety profile was similar to that of the first 6 months, with no new safety signals reported.

The efficacy and safety of daridorexant were evaluated in two 3-month-long studies, followed by either a 23-day safety follow-up period or enrollment in a 9-month, double-blind, placebo-controlled extension study. Data are presently available only up to 3 months for efficacy and 4 months for safety. Daridorexant 25 mg and 50 mg improved sleep outcomes, and daridorexant 50 mg also improved daytime functioning; both doses exhibited a favorable safety profile (see the Orexin Receptor Antagonists section above).

Despite favorable long-term findings, clinical wisdom suggests that hypnotics should be utilized for short periods of time as much as possible, and patients should be periodically evaluated during longer-term use. Withdrawal symptoms can occur in this class of compounds, and patients should be carefully monitored following abrupt discontinuation, especially in the case of agents with shorter half-lives. Even with medications prone to have these effects, the risk of rebound insomnia and withdrawal symptoms can be minimized by utilizing the lowest effective dose and by gradually tapering the dose downward over time for discontinuation.

PRN and Middle-of-the-Night Use

Insomnia is not a nightly occurrence in many patients. One study indicated that insomnia occurs on an intermittent basis in nearly half of all patients with insomnia seen in a primary care office. In this study, individuals with intermittent insomnia reported sleep problems an average of 4.6 nights per month, with each episode lasting 2.4 days. Insomnia was a problem an average of 5.1 days per month in the chronic insomnia group, and episodes lasted 15.3 days. Therefore, use of a medication for insomnia on an as-needed basis may be appropriate for many patients.

However, individuals whose insomnia is characterized by mid-nocturnal awakenings cannot utilize this approach with the majority of available hypnotic medications since MOTN administration, following awakenings, would introduce the possibility of next-day residual effects. In the case of the longer half-life medications, as-needed use is less possible; rather, prophylactic administration in the beginning of the night is necessary to avoid next-day carryover effects. Some have claimed that as-needed use is ideally suited for the treatment of a disorder such as insomnia that has an episodic longitudinal course, so that the medication can be taken only during an exacerbation of symptoms, thus diminishing the need for hypnotic medications. However, this contention has never been subjected to empirical validation.

Zaleplon’s short half-life of 1 hour makes it suitable for sleep induction, not maintenance. Nevertheless, in the case of discontinuous sleep, it has been shown to be effective in assisting patients in falling back to sleep if administered after awakenings in the middle of the night, although it is not specifically indicated for this usage by the FDA. If utilized in this fashion, the patient should be advised to remain in bed for a minimum of 4 hours after taking the medication to avoid daytime sedation. Zaleplon’s short half-life also increases flexibility regarding the timing of administration for patients who prefer to wait until the last minute, ie, until the need for a hypnotic becomes clearly apparent on any given night. This may make it more conducive for as-needed use, since patients would take it only if the need for a hypnotic persisted following a period of time in bed. Since zaleplon produces a short, 4-hour duration of impairment, it is also ideally suited for patients who have reduced sleep times and who have a need to be alert during the day.

The low-dose sublingual formulation of zolpidem was approved by the FDA specifically for the treatment of patients with insomnia characterized by difficulty falling asleep after awakening in the middle of the night. Please see more detailed discussion of sublingual zolpidem under the section on benzodiazepine receptor agonists.

Safety

As discussed, the main adverse effects of all of the hypnotic agents are daytime sedation and psychomotor and cognitive impairment. Essentially, these are products of the continuation of the drug’s activity during the day. The clinical challenge in using this class of compounds is the identification of a particular medication along the half-life continuum that will produce maximum efficacy across the night and minimum daytime residual effects.

Hypnotics should be used with caution in individuals with respiratory depression (e.g., COPD and OSAS), in elderly individuals, and in those with hepatic disease, those with multiple medical conditions, and those who are taking other medications that have CNS-depressant properties. They should not be used in pregnant women. Individuals who must awaken during the course of the drug’s active period should not take these medications. Table 11-7 outlines different populations that are vulnerable to the use of hypnotic medication.

The FDA has requested manufacturers of all sedative-hypnotic medications to include stronger language concerning the potential for severe allergic reactions and complex sleep-related behaviors, which may include sleep driving. Sleep driving is defined as driving while not fully awake after ingestion of a sedative-hypnotic product, with no memory of the event. The concern was triggered by anecdotal post-marketing reports of such events with some hypnotic agents. In the case of sleep driving, at least some of the episodes were associated with concomitant ingestion of alcohol and other sedating substances. It may be useful, therefore, to advise patients who are receiving hypnotic agents to limit the use of such substances whenever possible.

The FDA also required the lowering of recommended dosages for various zolpidem preparations including zolpidem, zolpidem ER and zolpidem SL due to the finding of high drug levels in the blood of some patients the day following drug administration that may impair activities that require alertness during the day, such as driving. These changes are reflected in Table 11-3. Women appear to be more susceptible to this than men, due to the slower rate of clearance of zolpidem in a small proportion of women.

For zolpidem and zolpidem SL, recommendations include the use of 5 mg for women, and 5 mg to 10 mg for men; for zolpidem ER, recommendations include the use of 6.25 mg for women and 6.25 mg or 12.5 mg for men. Recommended doses are also lower in women than in men in the case of low-dose zolpidem SL (1.75 mg vs 3.5 mg, respectively). Precautions regarding dosage for eszopiclone were set forth by the FDA due to the observation of next-day impairment at higher doses. Current recommendations include a starting dose of 1 mg at bedtime with the potential for an increase in dose, if clinically indicated, to a maximum of 3 mg. The dosing recommendations do not differ between men and women for eszopiclone.

For geriatric or debilitated patients, dose should not exceed 2 mg. For suvorexant, objective measures of next-day performance, including driving, indicated suvorexant was not associated with impairment for most patients. Assessment of driving in outpatient setting shows incidence of accidents and violations was low and comparable across treatments. Results did not differ by age subgroup. Generally, it is recommended that patients take the lowest necessary doses of hypnotic agents to minimize the effects of next-day impairment.

Dependence, Abuse and Other Precautions

Dependence (addiction) and abuse (an exaggerated desire to obtain the medication in increasing amounts to the exclusion of all other activities) continue to be significant concerns for physicians utilizing hypnotic agents. Figure 11-23 reviews the relative abuse liability of 19 hypnotic agents. Here, abuse liability is regarded as being a function of both the likelihood that a drug will be abused (used for nonmedical reasons) and the liability of abuse (ie, the untoward or toxic effects of using the drug nonmedically).

The first of these factors, the likelihood of abuse (Figure 11-23, yellow bars), is in turn determined by three factors, namely:

• The degree to which a compound functions as a reinforcer in drug self-administration studies conducted in nonhuman primates
• The extent of drug reinforcement and/or subjective drug liking in humans, as assessed by:
  • Prospective double-blind studies conducted in subjects with histories of drug abuse and assessing drug self-administration, drug choice, or ratings of liking/disliking or positive/negative subjective effects
  • Retrospective questionnaire studies of drug abusers and drug abuse clinicians who rate relative liking or preference for hypnotics based on abusers’ histories of exposure to these compounds
• The extent of actual abuse, an estimate of the relative rate of nonmedical use and recreational abuse of the individual hypnotics based on epidemiologic survey data and on case reports of abuse in the medical literature.
• The second of these factors, toxicity (Figure 11-23, green bars), is determined by:
  • The withdrawal severity after termination of chronic supratherapeutic doses of the drug
  • The degree of behavioral or cognitive impairment after acute administration of supratherapeutic doses
  • The likelihood of death after overdose.

These data suggest that abuse liability of hypnotics is highest for the barbiturate and barbiturate-like medications, intermediate for the BzRAs, low for trazodone and not present for ramelteon. Relative abuse liabilities for newer hypnotics were not assessed in this study.

Since the risk of abuse or problematic use of hypnotic drugs appears to be more likely in patients with histories of drug or alcohol abuse or dependence, hypnotics that carry a DEA schedule should be used in caution or not used at all in patients with such backgrounds. Other groups at risk for the development of problematic hypnotic use include the elderly and patients with chronic pain.

Chronic use of hypnotics has been regarded by many as a measure of abuse. In fact, case reports of physical dependence at appropriate doses of hypnotics following chronic use exist, yet long-term studies examining this question are lacking. Epidemiologic data suggest that the majority of patients use hypnotics for ≤2 weeks and that those who utilize them on a chronic basis do not frequently display dose escalation. Rather, the extent of hypnotic self-administration seems to depend mainly upon the severity of the sleep disturbance on the prior night, indicating a therapeutic pattern of use. Nevertheless, a pattern of long-term use with dose escalation can indicate the potential for hypnotic abuse and such patients should be monitored closely.

In contrast to BzRAs and suvorexant, doxepin and ramelteon are not associated with abuse potential, nor do they appear to produce physical dependence. Nevertheless, patients should be monitored for signs of drug abuse, and observed carefully.

Abuse of suvorexant poses an increased risk of somnolence, daytime sleepiness, decreased reaction time, and impaired driving skills. Patients at risk for abuse may include those with prolonged use of suvorexant, those with a history of drug abuse, and those who use suvorexant in combination with alcohol or other abused drugs. In an abuse liability study conducted in recreational polydrug users (n=36), suvorexant (40, 80, and 150 mg) produced similar effects as zolpidem (15, 30 mg) on subjective ratings of “drug liking” and other measures of subjective drug effects.

The abuse potential of lemborexant was assessed in a study of 29 recreational sedative abusers. The “drug liking”, “overall drug liking”, “take drug again”, and “good drug effects” ratings for lemborexant 10 mg, 20 mg and 30 mg were statistically similar to zolpidem 30 mg and suvorexant 40 mg, and higher than those for placebo.

The abuse potential of daridorexant was evaluated in 63 recreational sedative drug users. Compared to zolpidem 30 mg and suvorexant 150 mg, daridorexant demonstrated considerably lower “drug liking” ratings at the dose of 50 mg, but significantly higher “drug liking” ratings than placebo. Daridorexant had similar “drug liking” ratings to zolpidem 30 mg and suvorexant 150 mg at dosages of 100 mg and 150 mg (two and three times the maximum recommended dose, respectively). No reports suggestive of abuse liability were made in the placebo-controlled phase 3 clinical studies of daridorexant, in which patients received the drug for up to a year.

Because individuals with a history of abuse or addiction to alcohol or other drugs may be at increased risk for abuse and addiction to suvorexant, lemborexant and daridorexant, these patients must be followed carefully.

General guidelines for prescribing hypnotic medications, listing factors that should be taken into consideration, are presented in Table 11-8.

FDA-Unapproved Agents for Insomnia

Agents in this class are used as hypnotics but not indicated for this use by the FDA. Factors favoring their use include low abuse liability, availability of wide dose ranges and, in some cases, low cost. Although these agents may have been studied for their effects on sleep in other conditions complicated by insomnia, only information relative to primary insomnia will be reviewed here.

Sedating Antidepressants

These agents are prescribed extensively despite limited data on their safety and efficacy in insomnia. For insomnia, they are typically utilized at doses that are subtherapeutic for the treatment of depression or anxiety disorders.

Trazodone is a heterocyclic antidepressant that has an elimination half-life of 5 to 12 hours. It has received little scientific attention as a sleep aid in primary insomnia (reviewed by James and Mendelson). Walsh and colleagues reported on a 14-day controlled investigation of trazodone 50 mg, zolpidem 10 mg and placebo in patients with primary insomnia. During the first week, both active drugs reduced subjective sleep latency, although trazodone was significantly less effective than zolpidem. By the second week, trazodone’s effects were no different from those of placebo, yet the dosage of trazodone was not increased to assess the potential for maintenance of efficacy at higher dosages.

Doxepin is a tricyclic antidepressant (TCA). At doses of 25 mg to 50 mg, it was the subject of a controlled study in 47 subjects with primary insomnia for 4 weeks; it demonstrated an improvement in TST but not in sleep latency. Withdrawal insomnia was evident following abrupt discontinuation. As reviewed above, it is approved for the treatment of insomnia in lower dosages of 3 mg and 6 mg. Mirtazapine is a newer antidepressant with an elimination half-life of 20 to 40 hours. Although mirtazapine has been shown to improve symptoms of insomnia associated with major depressive disorder, there are no published reports on its use in primary insomnia patients.

Therefore, despite their potential advantages, the paucity of available data regarding their effects on sleep and wakefulness in insomnia limits the use of these agents in the management of insomnia. Their use is also complicated by daytime sedation and cognitive impairment, anticholinergic effects, weight gain and drug-drug interactions. The TCAs are also potentially fatal in an overdose.

Antiepileptic Drug

The only antiepileptic agent that has been the subject of a formal large-scale investigation for primary insomnia is tiagabine. However, at doses ranging from 2 to 8 mg over the course of two nights, it did not have a significant effect on WASO, sleep latency, TST, or the subjective rating of sleep. Its main adverse effects included residual daytime somnolence at the highest dose of 8 mg.

Atypical Antipsychotics

Several sedating antipsychotics (e.g., quetiapine and olanzepine) are occasionally used alone for the treatment of insomnia, even though they are not approved for this use. Nevertheless, studies demonstrating the usefulness for the management of insomnia are scant and suffer from methodological drawbacks. A recent literature review of the published studies and case reports in which quetiapine was used specifically for the treatment of insomnia as the primary endpoint concluded that while it has moderate sedative properties and was reported to provide improvements in several subjective and objective sleep parameters, it, as well as other antipsychotics, can have adverse effects, such as periodic limb movements (PLMs), akathisia, metabolic complications, and weight gain. Therefore, although the use of antipsychotics for various psychiatric disorders that feature insomnia may be appropriate for the treatment of the primary psychiatric condition, their use in the treatment of primary insomnia cannot be recommended.

Clinical Practice Guidelines

The American Academy of Sleep Medicine (AASM) commissioned a task force of four experts in sleep medicine to develop recommendations for pharmacological treatment of insomnia disorder. To determine the direction and strength of a recommendation, the task force assessed quality of evidence, balance of beneficial and harmful effects and patient values and preferences. The GRADE approach (Grades of Recommendation, Assessment, Development and Evaluation) was used to assess the quality of evidence. A STRONG recommendation under GRADE is one that clinicians should generally abide by. A WEAK recommendation does not necessarily indicate ineffectiveness; rather, it represents a reduced level of confidence in the outcome and appropriateness of the patient-care plan for all patients.

The 2017 AASM guideline is based on a systematic review (including meta-analyses) of pharmacological agents used for the management of insomnia and provides recommendations for orexin receptor antagonists, BzRAs and benzodiazepines, melatonin agonists, heterocyclics, anticonvulsants and over-the-counter preparations. The guideline also includes literature reviews of estazolam, quazepam, flurazepam, oxazepam, quetiapine, gabapentin, paroxetine and trimipramine; for these agents, clinical practice recommendations were not possible due to inadequate data for statistical analyses. The 2017 AASM recommendations are summarized in Table 11-9.

Importantly, the 2017 AASM guideline states that it should be used in combination with previous AASM guidelines on the evaluation and management of chronic insomnia, as it is intended to act as only one element in an ongoing assessment of a patient with insomnia. Another important statement from the guideline is that medications for chronic insomnia disorder should be considered mainly in patients who are unable to participate in CBT-I, who still have symptoms despite participation in such treatments, or, in select cases, as a temporary adjunct to CBT-I.

In 2023, the Alliance for Sleep, a committee of 5 US-based sleep medicine experts published a guideline on switching or deprescribing hypnotic medications for insomnia. Briefly, this guideline recommends tapering for the following drugs: benzodiazepines when switching to a different drug class (but not to another benzodiazepine), zolpidem and eszopiclone when switching to another Z-drug or another drug class (not required for zaleplon) and trazodone, mirtazapine, TCAs and quetiapine when switching to a different drug class (within-class switching is not recommended for these drugs). A direct switch is recommended for orexin receptor antagonists, ramelteon, or dopexin.