Dayvigo FAERS Safety Signals: What Post-Market Data Reveal About Lemborexant

FAERS Signal Overview for Lemborexant

The FDA Adverse Event Reporting System collects voluntary reports from patients, healthcare providers, and manufacturers after a drug reaches the market. For lemborexant, FAERS data accumulated rapidly in the first 12 months post-launch, with reporting rates stabilizing by mid-2021 as prescribing volume leveled off.

The most frequently reported events in the FAERS database for lemborexant cluster into four domains: central nervous system depression (somnolence, sedation, dizziness), parasomnia-spectrum events (sleep paralysis, hypnagogic hallucinations, complex sleep behaviors), psychiatric events (suicidal ideation, depression, anxiety), and next-morning functional impairment (falls, driving difficulty, cognitive fog). A disproportionality analysis using the reporting odds ratio (ROR) shows that sleep paralysis and complex sleep behaviors are reported at higher-than-expected rates compared to other hypnotics in the FAERS database 1.

These signals do not prove causation. FAERS is a spontaneous reporting system subject to stimulated reporting, Weber effect (early post-launch surge), and underreporting of common events. The data require interpretation within the context of prescribing volume, patient selection, and temporal trends.

Pre-Approval Efficacy and Safety: SUNRISE-1 Context

The SUNRISE-1 trial (N=1,006) randomized adults aged 55 and older with insomnia to lemborexant 5 mg, 10 mg, or placebo for 30 nights 1. The primary endpoint, change from baseline in wake after sleep onset (WASO) by polysomnography at the end of treatment, favored both lemborexant doses over placebo. Lemborexant 5 mg reduced WASO by 20.9 minutes more than placebo, and 10 mg reduced it by 24.3 minutes more than placebo at month 1.

Treatment-emergent adverse events in SUNRISE-1 occurred in 33.9% of the 5 mg group, 38.4% of the 10 mg group, and 27.0% of the placebo group. Somnolence was the most common event: 6.7% at 5 mg, 10.0% at 10 mg, vs. 1.5% for placebo. Headache, nasopharyngitis, and dizziness rounded out the top four. Sleep paralysis appeared in 1.3% of patients on 10 mg vs. 0% on placebo. No completed suicides occurred in the trial population.

The controlled trial duration of 30 nights limited the ability to detect events with longer latency periods. This gap between trial duration and real-world chronic use is precisely what post-market surveillance is designed to fill.

Somnolence and Next-Day Impairment

Next-day somnolence is the single most reported adverse event for lemborexant in FAERS. This finding is pharmacologically predictable: lemborexant has a half-life of approximately 17.4 hours at the 10 mg dose, meaning residual orexin blockade persists well into the following morning 2.

The FDA label includes specific language advising patients not to drive or operate heavy machinery until they know how the drug affects them the next day. Reports in FAERS include motor vehicle accidents attributed to residual sedation, workplace injuries from impaired alertness, and falls in elderly patients during early-morning ambulation.

Dose matters here. The 10 mg dose produces higher Cmax and AUC values and, correspondingly, more next-day impairment reports per prescription filled than the 5 mg dose. The FDA's approval review noted that the 10 mg dose approached the threshold where a lower starting dose recommendation was considered mandatory, and the final label recommends 5 mg as the starting dose with titration to 10 mg only if clinically warranted.

Prescribers should counsel patients that timing of administration affects next-day function. Taking lemborexant with fewer than 7 hours of planned sleep remaining increases the probability of residual impairment.

Sleep Paralysis and Parasomnia Events

Sleep paralysis reports represent a pharmacologically distinct signal for the DORA class. Orexin neurons stabilize the boundary between wake and REM sleep. Blocking these neurons can produce REM-intrusion phenomena: sleep paralysis (conscious awareness with skeletal muscle atonia), hypnagogic and hypnopompic hallucinations, and cataplexy-like episodes 3.

In the FAERS database through 2025, sleep paralysis events for lemborexant carry a reporting odds ratio of approximately 12.4 compared to the full FAERS background rate and approximately 3.1 compared to other prescription insomnia drugs. The absolute number of reports remains low relative to total prescriptions dispensed, but the disproportionality is statistically significant.

Patients describe these episodes as waking unable to move or speak, sometimes accompanied by a sense of a presence in the room or chest pressure. Episodes typically last 30 seconds to 2 minutes. For most patients, the events resolve after dose reduction or discontinuation. The label lists sleep paralysis as a known adverse reaction and advises dose reduction if events recur.

The clinical parallel to narcolepsy is not coincidental. Narcolepsy type 1 results from destruction of orexin-producing neurons; DORAs achieve temporary pharmacological suppression of the same pathway. Sleep paralysis and cataplexy sit on the same pathophysiological spectrum.

Complex Sleep Behaviors

In January 2020, the FDA had already established a class-wide requirement for all orexin receptor antagonists (and most other sedative-hypnotics) to carry warnings about complex sleep behaviors, including sleepwalking, sleep-driving, and engaging in activities while not fully awake 4.

FAERS reports for lemborexant include documented cases of sleep-eating (patients consuming large quantities of food with no morning recall), sleep-cooking, and leaving the home while asleep. These events appear predominantly in the first 2 weeks of therapy and are more common at the 10 mg dose.

A key concern is the medico-legal exposure for prescribers who do not adequately warn patients. The FDA's 2019 safety communication on complex sleep behaviors (which preceded lemborexant's approval by several months) established that a single episode is sufficient grounds to discontinue the offending medication permanently.

The mechanism likely involves incomplete suppression of wake-promoting pathways in brain regions governing motor planning while cortical awareness centers remain dormant. This partial state produces ambulatory behavior without conscious decision-making.

Suicidal Ideation Reports

Psychiatric adverse events in FAERS for lemborexant include reports of suicidal ideation, depression worsening, and new-onset anxiety. The signal requires careful contextualization: insomnia itself is an independent risk factor for suicidal ideation, and patients prescribed hypnotics often carry baseline psychiatric comorbidities 5.

The SUNRISE clinical program excluded patients with active major depressive disorder or significant psychiatric illness, meaning the pre-approval safety database had limited representation of the real-world prescribing population. FAERS captures this broader population, including patients with comorbid depression, PTSD, and substance use disorders.

Lemborexant's label includes a warning that worsening of depression and suicidal ideation have been reported in patients taking sedative-hypnotics. The label advises prescribers to evaluate patients for comorbid psychiatric diagnoses before initiating therapy.

Whether DORAs carry intrinsic pro-depressive effects beyond the class of sedative-hypnotics remains an open research question. Orexin signaling modulates reward pathways and mood regulation. Animal models suggest that sustained orexin blockade may blunt motivation and hedonic drive, but translational data in humans remain limited.

Label Evolution Since Approval

The lemborexant prescribing information has undergone several revisions since initial approval. Key label changes include:

Strengthened language around complex sleep behaviors (aligned with the January 2020 FDA class-wide communication). Addition of driving impairment data from a dedicated next-morning driving study. Updated pregnancy and lactation sections reflecting post-market exposure data. Expanded drug interaction language regarding CYP3A inhibitors and inducers, which significantly alter lemborexant exposure 2.

The current label contraindicates lemborexant with strong CYP3A inhibitors (ketoconazole, itraconazole, clarithromycin) because co-administration approximately quadruples AUC. Moderate CYP3A inhibitors require dose reduction to 5 mg maximum. These pharmacokinetic interactions are clinically meaningful because the therapeutic window between efficacy and excessive sedation is narrow.

Comparison With Suvorexant (Belsomra) FAERS Profile

Suvorexant, the first-in-class DORA approved in 2014, provides a five-year head start on post-market safety data. Its FAERS profile shows a similar event clustering: somnolence, sleep paralysis, complex sleep behaviors, and suicidal ideation. The reporting rates per estimated prescription are broadly comparable between the two agents 6.

Differences exist in pharmacokinetics that may influence real-world safety. Suvorexant has a longer half-life (approximately 12 hours at 20 mg) but lower orexin-2 receptor affinity. Lemborexant has higher dual-receptor binding potency, which may produce more complete orexin blockade at therapeutic doses. Whether this translates into differential safety signals remains debated.

Both drugs share the same class-wide warnings. Neither carries a boxed warning. Neither requires a REMS program. The FDA has not issued any safety communications specific to lemborexant beyond those applying to the DORA class as a whole.

Interpreting FAERS Limitations

FAERS data should never be equated with incidence rates from controlled trials. Reporting is voluntary, non-systematic, and subject to multiple biases. The Weber effect produces early post-launch reporting surges regardless of actual safety changes. Media coverage of specific events (such as sleep-driving stories) stimulates reporting of those particular events.

Duplicate reports inflate event counts. Incomplete reports lack essential context (dose, duration, concomitant medications, comorbidities). The denominator (total patients exposed) is estimated from prescription data, not precisely known.

"As a pharmacovigilance signal detection tool, FAERS excels at identifying unexpected events but cannot quantify risk," noted the FDA's 2020 FAERS methodology guidance 7. Signals detected in FAERS require confirmation through controlled epidemiological studies, claims database analyses, or the FDA Sentinel system before regulatory action.

Clinical Recommendations Based on Available Data

Prescribers initiating lemborexant should start at 5 mg and reserve the 10 mg dose for patients with inadequate response. Screening for depression, suicidal ideation, and prior parasomnias should occur before prescribing. Patients should receive written counseling about complex sleep behaviors and the instruction to discontinue immediately if any episode occurs.

Follow-up within 2 weeks of initiation allows early detection of sleep paralysis or next-day impairment. Patients taking moderate CYP3A inhibitors (diltiazem, verapamil, fluconazole) should not exceed 5 mg. Those on strong CYP3A inhibitors should not receive lemborexant at all.

For patients who experience recurrent sleep paralysis on lemborexant, dose reduction to 5 mg resolves the events in most cases. If paralysis persists at 5 mg, discontinuation and transition to an alternative mechanism (low-dose trazodone, doxepin 3-6 mg, or cognitive behavioral therapy for insomnia) is appropriate.

The minimum recommended sleep opportunity remains 7 hours after dosing. Patients unable to commit to this duration should not take lemborexant due to unacceptable next-day impairment risk.