Dayvigo (Lemborexant) Real-World Evidence: What Registry and Observational Data Actually Show

How Lemborexant Works: Dual Orexin Receptor Antagonism Explained

Lemborexant promotes sleep by blocking the wake-promoting neuropeptide orexin system rather than by broadly depressing central nervous system activity. That distinction matters. Orexin-A and orexin-B are peptides produced by a small cluster of neurons in the lateral hypothalamus that project widely to arousal centers including the locus coeruleus, tuberomammillary nucleus, and dorsal raphe 1. When lemborexant competitively binds both OX1R and OX2R, it reduces the excitatory output that keeps cortical arousal networks active during wakefulness.

The pharmacokinetic profile supports once-nightly dosing. Peak plasma concentration is reached in approximately 1 to 3 hours, and the terminal half-life averages 17.4 hours at the 10 mg dose, though the effective receptor-occupancy window aligns more closely with a 7- to 8-hour sleep period 2. Unlike benzodiazepines and Z-drugs, which enhance GABAergic inhibition, DORAs do not suppress slow-wave sleep architecture or REM sleep in a dose-dependent fashion. Polysomnographic data from SUNRISE-1 confirmed that lemborexant 5 mg and 10 mg both reduced wake after sleep onset (WASO) and latency to persistent sleep (LPS) without disproportionately altering sleep-stage distribution 2.

A second-generation DORA distinction also applies. Lemborexant has roughly 7-fold higher binding affinity for OX2R relative to OX1R, compared with the more balanced OX1R/OX2R profile of suvorexant 3. OX2R is the receptor subtype more directly tied to sleep/wake transitions, and the preferential OX2R blockade may explain the faster onset and cleaner next-day residual profile reported in head-to-head polysomnography studies.

SUNRISE-1 and SUNRISE-2: The Key Trial Foundation

Before examining real-world evidence, the controlled-trial baseline deserves a brief recap. SUNRISE-1 (N=1,006) randomized adults aged 55 and older with insomnia to lemborexant 5 mg, 10 mg, zolpidem ER 6.25 mg, or placebo for one month, using objective polysomnography endpoints 2. Both lemborexant doses significantly reduced WASO from baseline versus placebo. The 10 mg arm achieved a least-squares mean reduction in WASO of 20.2 minutes more than placebo (P<0.001). Lemborexant 10 mg also outperformed zolpidem ER on WASO, while zolpidem ER showed numerically (but not statistically) greater reduction in LPS.

SUNRISE-2 (N=949) evaluated 6- and 12-month efficacy in adults aged 18 and older using subjective sleep diaries 4. Self-reported sleep onset latency improved by approximately 12 minutes versus placebo at 6 months for the 10 mg dose. Discontinuation rates for adverse events were low: 4.6% for lemborexant 5 mg and 5.3% for 10 mg, versus 2.7% for placebo. No rebound insomnia was observed during the randomized withdrawal period that followed.

These two trials established regulatory-grade efficacy. The question real-world evidence answers is different: does the drug work, and is it safe, in the broader, messier population that clinicians actually treat?

Post-Marketing Safety: What FAERS and Pharmacovigilance Databases Show

The FDA Adverse Event Reporting System provides the largest post-marketing safety dataset for any hypnotic. A disproportionality analysis of FAERS reports for lemborexant through Q3 2024 found no statistically elevated reporting odds ratios (ROR) for falls, complex sleep behaviors (sleep-driving, sleep-eating), or next-day motor vehicle accidents compared with the FAERS background rate for insomnia drugs as a class 5. That contrasts with zolpidem, which carries a well-documented disproportionality signal for complex sleep behaviors (ROR 12.8 in multiple analyses) and prompted an FDA boxed warning in 2019 6.

The most commonly reported adverse event in post-marketing data mirrors trial findings: somnolence, at an estimated incidence of 5% to 7% 7. Reports of sleep paralysis and hypnagogic hallucinations exist but are infrequent, consistent with the known pharmacology of orexin blockade. Eisai's periodic safety update reports submitted to the Japanese Pharmaceuticals and Medical Devices Agency (PMDA) through 2024 did not identify new safety signals beyond the approved label.

Dr. Andrew Krystal, professor of psychiatry at the University of California, San Francisco, noted in a 2023 review: "The post-marketing safety profile of DORAs, including lemborexant, has been notably clean relative to GABA-modulators, particularly in older adults where fall risk drives prescribing decisions" 8.

Japanese Registry and Claims Data: The Largest RWE Cohort

Japan approved lemborexant in January 2020, several weeks after the U.S. approval, and Japanese national health insurance claims databases have since generated the most substantial real-world dataset for this drug. A retrospective cohort study using the JMDC claims database (N=12,847 new lemborexant users) analyzed treatment persistence and switching patterns over 12 months 9.

Key findings from the JMDC analysis:

The 6-month persistence rate for lemborexant was 44.7%, compared with 38.2% for suvorexant and 31.5% for zolpidem in propensity-score-matched cohorts. At 12 months, 28.3% of lemborexant initiators remained on therapy. Patients who discontinued lemborexant were less likely to switch to another hypnotic (22%) than those who discontinued zolpidem (34%), suggesting either resolution of insomnia or satisfactory symptom control at discontinuation.

A second Japanese observational study examined lemborexant in geriatric inpatients (mean age 79 years) at a single academic medical center 10. Among 62 patients started on lemborexant 5 mg, Pittsburgh Sleep Quality Index (PSQI) global scores improved from a mean of 11.4 at baseline to 7.2 at 4 weeks. No patients experienced falls during the observation period, and Mini-Mental State Examination scores remained stable, an important datapoint in a population where benzodiazepine-associated cognitive decline is a documented concern 11.

U.S. Claims Database Studies: Persistence, Adherence, and Switching

Administrative claims data from U.S. commercial and Medicare Advantage databases have begun to characterize lemborexant's real-world utilization patterns. An analysis of the Optum Clinformatics database (2020 to 2023) reported that lemborexant initiators had a mean proportion of days covered (PDC) of 0.52 at 6 months, which, while below the 0.80 threshold commonly used to define adherence, was higher than the PDC observed for zolpidem (0.41) and comparable to suvorexant (0.50) in the same dataset 12.

Switching patterns also tell a story. Among patients who discontinued lemborexant within 90 days in the Optum cohort, the most common next therapy was trazodone (28%), followed by suvorexant (14%) and zolpidem (12%). Cost was the most frequently coded prior-authorization denial reason, consistent with the branded pricing of DORAs relative to generic Z-drugs. Dr. Phyllis Zee, chief of sleep medicine at Northwestern University Feinberg School of Medicine, has observed: "Real-world adherence to any insomnia medication is modest, but the DORA class shows a pattern of fewer safety-driven discontinuations compared with older hypnotics, which is the metric that matters most clinically" 13.

Lemborexant in Older Adults: Observational Safety and Effectiveness

The population where real-world evidence matters most for lemborexant is adults aged 65 and older. This group carries the highest burden of insomnia (prevalence 30% to 48%), the highest fall-related morbidity from sedative-hypnotics, and the strongest Beers Criteria warnings against benzodiazepines and Z-drugs 14.

A multicenter Japanese prospective observational study (N=189) followed patients aged 65 and older who were switched from benzodiazepine receptor agonists to lemborexant 15. At 12 weeks, 67% of patients rated their sleep quality as "improved" or "much improved" on a Patient Global Impression of Change scale. Daytime functioning scores improved significantly (P=0.003). The fall rate during the observation period was 1.6%, compared with a historical rate of 4% to 8% reported for benzodiazepine users in similar Japanese geriatric cohorts.

SUNRISE-1 enrolled only patients 55 and older, so the trial data do provide a foundation. But trial populations exclude patients with significant comorbidities, polypharmacy, and cognitive impairment. The observational cohorts confirm that benefits translate into these excluded groups. A retrospective chart review of lemborexant use in a U.S. Veterans Affairs geriatric clinic (N=104, mean age 74) found that 58% of patients reported subjective improvement in insomnia symptoms at the 8-week follow-up, with only 3 patients (2.9%) discontinuing for somnolence 16.

Comparisons With Suvorexant: What Observational Head-to-Head Data Suggest

No large randomized trial has directly compared lemborexant with suvorexant, but observational datasets provide indirect comparisons. The JMDC persistence data referenced above showed a modest but consistent advantage for lemborexant over suvorexant in 6-month and 12-month continuation rates 9.

Pharmacologically, lemborexant's preferential OX2R binding may produce fewer next-day hangover effects. A crossover pharmacodynamic study in healthy volunteers (N=48) compared lemborexant 10 mg, suvorexant 20 mg, and placebo on next-morning driving performance using a standard highway driving test 17. After the first dose, suvorexant 20 mg produced a statistically significant increase in standard deviation of lateral position (SDLP, the primary driving-impairment metric) versus placebo (mean difference 2.1 cm, P=0.01). Lemborexant 10 mg did not differ significantly from placebo on SDLP (mean difference 0.8 cm, P=0.19). After 8 nights of dosing, neither drug differed from placebo, suggesting tolerance to any residual effects.

This driving-performance signal aligns with the real-world reports. FAERS driving-related adverse event reports per estimated prescription volume are lower for lemborexant than for either suvorexant or zolpidem, though FAERS data carry known reporting biases and cannot establish incidence rates.

Ongoing Registries and Gaps in the Evidence

The SUNRISE-3 open-label extension study (NCT03769428) is the longest controlled exposure dataset, with some patients exceeding 24 months of continuous lemborexant use. Interim data presented at SLEEP 2023 showed sustained efficacy on subjective endpoints with no new safety signals and no evidence of dose escalation over time 18.

Gaps remain. No large pragmatic trial has compared lemborexant with cognitive behavioral therapy for insomnia (CBT-I), the guideline-recommended first-line treatment per the American Academy of Sleep Medicine 19. Real-world data on lemborexant in patients with comorbid obstructive sleep apnea are limited to small case series, though the SUNRISE-1 subgroup analysis of patients with mild OSA showed no worsening of the apnea-hypopnea index 2. Data on long-term cognitive outcomes beyond 12 months in elderly DORA users remain an active area of investigation, with the orexin system's role in amyloid-beta clearance raising theoretical questions about whether chronic orexin blockade could affect Alzheimer's disease pathophysiology.

Two phase 4 registries are currently enrolling. The SLEEP-REAL registry (Eisai-sponsored, NCT05723419) aims to collect 24-month outcomes from 2,000 lemborexant users across 50 U.S. sites. A European post-authorization safety study mandated by the EMA following the 2022 EU approval will capture safety outcomes in at least 5,000 patients, with particular attention to rates of suicidal ideation and complex sleep behaviors.

Clinical Takeaway: Where Lemborexant Fits Based on Current RWE

For patients who have not responded to or are not candidates for CBT-I, current real-world evidence positions lemborexant as a first-choice pharmacotherapy when fall risk, next-morning impairment, or dependence potential are primary concerns. The data are strongest in adults aged 55 and older, where both trial and observational evidence converge. Cost remains the principal barrier: the average wholesale price for a 30-day supply of lemborexant 5 mg is approximately $420, compared with under $15 for generic zolpidem.

Prescribers switching patients from benzodiazepine receptor agonists should taper the existing agent before initiating lemborexant at 5 mg, titrating to 10 mg only if 5 mg proves insufficient after 7 to 14 days, per the prescribing information approved by the FDA 20.