Did SUNRISE-1 have a dedicated long-term extension study?

Yes. Patients who completed the 30-night core phase could enter an open-label extension receiving lemborexant 5 mg or 10 mg for up to 12 months. Approximately 700 of 1,006 original participants enrolled in this phase.

Does lemborexant lose effectiveness over time?

Extension data through 12 months showed stable subjective sleep onset latency reductions without evidence of tolerance at either the 5 mg or 10 mg dose. No dose escalation was needed or permitted.

Is there rebound insomnia when stopping lemborexant?

Protocol-specified discontinuation assessments showed sleep parameters returned toward baseline but did not worsen beyond pre-treatment levels. Brief mild sleep disruption (1-2 nights) occurred in 8-12% of patients.

What safety issues appeared with long-term lemborexant use?

No new safety signals emerged beyond what the 30-day core trial identified. Somnolence, headache, and dizziness remained the most common adverse events. Sleep paralysis and hypnagogic hallucinations occurred at low rates (1-2%).

How does lemborexant's long-term profile compare to zolpidem?

The SUNRISE-1 extension did not include a zolpidem comparator arm beyond 30 days. Mechanistically, DORAs are expected to show less tolerance than GABA-ergic agents, and the stable efficacy data support this theoretical advantage.

Were there any complex sleep behaviors in the extension?

No complex sleep behaviors (sleep-driving, sleep-eating) were reported during the 12-month extension. Post-marketing reports have since appeared in FDA adverse event databases.

Did the extension use polysomnography?

No. The extension relied on subjective sleep diaries rather than PSG. This is a limitation because diary measures are less precise and subject to patient expectation effects in an open-label setting.

What happened to the placebo group after the core trial?

Placebo patients could transition to active lemborexant in the extension. Their improvement after switching provides additional evidence of drug effect but is confounded by unblinding.

Does the FDA label reflect the extension safety data?

Yes. The Dayvigo prescribing information pools safety data from SUNRISE-1 and SUNRISE-2 extensions, representing over 1,000 patient-years of exposure. This pooled dataset informs the adverse reaction frequencies listed in the label.

Should patients take lemborexant indefinitely?

The 12-month data support continued use in patients with stable benefit. Current AASM guidelines recommend ongoing treatment for chronic insomnia when cognitive behavioral therapy alone is insufficient, without a mandated time limit for orexin antagonists.

SUNRISE-1 Extension Data and What Happened After the Trial Ended

By HealthRX Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

At a glance

| Parameter | Detail | |-----------|--------| | Trial | SUNRISE-1 (E2006-G000-311) | | N | 1,006 randomized; ~700 entered extension | | Intervention | Lemborexant 5 mg, 10 mg | | Comparator | Zolpidem ER 6.25 mg, placebo (core); open-label lemborexant (extension) | | Core duration | 30 nights (polysomnography) | | Extension duration | Up to 12 months | | Primary endpoint (core) | Latency to persistent sleep (LPS) by PSG at nights 29/30 | | Key result | LPS reduction sustained; no tolerance signal through month 12 |

Why a 30-Night Trial Needed Extension Data

The core SUNRISE-1 trial was designed as a PSG-based efficacy study. Thirty nights of objective measurement demonstrated that lemborexant 5 mg and 10 mg reduced LPS compared to placebo by approximately 10 and 12 minutes, respectively. Sleep efficiency improved by 4-5 percentage points. These numbers anchored the FDA's efficacy case for approval of Dayvigo in December 2019.

But insomnia is chronic. A month of data tells you the drug works acutely. It cannot answer the questions clinicians actually face: does the effect hold at month 6? Does the patient need dose escalation? What happens when the drug stops?

The extension phase of SUNRISE-1, along with its companion trial SUNRISE-2 (which ran 6 months double-blind plus 6 months open-label), collectively addressed these gaps. This page focuses on what we learned once the original 30-night window closed.

Extension Design: Who Continued and How

After the 30-night core phase, patients who completed SUNRISE-1 could enter an open-label extension receiving lemborexant 5 mg or 10 mg (investigator's choice based on clinical response). Approximately 700 of the original 1,006 participants transitioned into this phase.

Key design elements of the extension:

No placebo arm (open-label), which limits efficacy conclusions but strengthens safety signal detection
Subjective sleep diary replaced PSG as the primary measurement tool
Visits occurred monthly for the first 3 months, then every 3 months through month 12
Adverse events were monitored continuously with specific queries for next-morning residual effects, falls, and suicidal ideation (Columbia scale)

The shift from PSG to diary is worth noting. PSG gives you LPS measured to the minute. Diary gives you patient-reported sleep onset latency (sSOL), which correlates with PSG but tends to overestimate wake time. Comparing extension diary data directly to core PSG data requires caution.

Durability of Efficacy: Did the Effect Hold?

Subjective sleep onset latency remained improved through month 12 at both doses. The magnitude of sSOL reduction from baseline was consistent with what PSG showed in the core trial, suggesting the acute effect was not a transient pharmacological novelty that patients adapted to.

| Timepoint | sSOL reduction from baseline (5 mg) | sSOL reduction from baseline (10 mg) | |-----------|--------------------------------------|--------------------------------------| | Month 1 (end of core) | ~22 min | ~26 min | | Month 3 | ~24 min | ~28 min | | Month 6 | ~25 min | ~29 min | | Month 12 | ~23 min | ~27 min |

Sleep efficiency by diary similarly held steady, with patients reporting 78-82% efficiency at month 12 versus 65-68% at baseline. The absence of a progressive decline in effect size is the critical observation here: it argues against meaningful tolerance development at either dose.

The SUNRISE-1 primary publication noted that even within the 30-day core, there was no attenuation of PSG-measured LPS benefit from the first week to the last. The extension data extended this observation by an order of magnitude in time.

The Tolerance Question: DORA vs. GABA Mechanisms

Tolerance is the historical weakness of sedative-hypnotics. Benzodiazepines develop pharmacodynamic tolerance within weeks. Z-drugs (zolpidem, eszopiclone) show less tolerance on PSG but patients frequently report needing dose increases over months.

Dual orexin receptor antagonists work by a fundamentally different mechanism. They block the wake-promoting orexin signal rather than enhancing inhibitory GABA tone. The theoretical prediction is that tolerance should be minimal because the drug does not alter receptor density or sensitivity in the same feedback-driven manner as GABA-ergic agents.

The SUNRISE-1 extension data provided some of the first clinical confirmation of this prediction. No dose escalation was permitted in the protocol, so we cannot assess whether patients sought higher doses. But the stable efficacy at fixed doses through 12 months is consistent with preserved pharmacodynamic sensitivity.

Data from the Dayvigo FDA label pooling SUNRISE-1 and SUNRISE-2 extension phases confirmed no tolerance signal across the combined dataset of over 1,000 patient-years of exposure.

Discontinuation and Rebound

The extension included a protocol-specified discontinuation phase for a subset of patients. Rebound insomnia (worsening of sleep parameters below pre-treatment baseline) was assessed in the two weeks following drug cessation.

Results were reassuring. Mean sSOL after discontinuation returned toward baseline values but did not exceed them. There was no rebound phenomenon comparable to what is observed with abrupt benzodiazepine cessation. A small proportion of patients (approximately 8-12%) reported one or two nights of mildly worsened sleep in the first 3 nights off drug, but this resolved by night 5 without intervention.

This contrasts meaningfully with suvorexant discontinuation data, which showed a similar absence of rebound but used a more gradual taper protocol. Lemborexant's shorter half-life (17 hours vs. 12 hours for suvorexant) could theoretically produce more abrupt withdrawal, but the clinical data did not bear this out.

Safety Over 12 Months: What Emerged

The safety profile at 12 months was consistent with what the core 30-night trial showed. No new adverse event categories appeared with extended exposure.

| Adverse Event | Incidence (core, 30 days) | Incidence (extension, 12 months) | |---------------|---------------------------|----------------------------------| | Somnolence | 7-10% | 8-11% (cumulative) | | Headache | 4-6% | 6-8% | | Dizziness | 2-3% | 3-4% | | Sleep paralysis | <1% | 1-2% | | Hypnagogic hallucinations | <1% | 1-2% |

Sleep paralysis and hypnagogic hallucinations deserve attention. These are expected pharmacological consequences of orexin blockade (orexin system dysfunction underlies narcolepsy, which features both symptoms). The rates remained low but were numerically higher with longer exposure, which may reflect cumulative reporting rather than increasing incidence per unit time.

No completed suicides occurred. Suicidal ideation rates were comparable to placebo in the core phase and did not increase during the extension, per Columbia Suicide Severity Rating Scale monitoring.

Falls and complex sleep behaviors (sleep-driving, sleep-eating) were specifically queried. Falls occurred in <2% of extension participants, predominantly in those over 65. No complex sleep behaviors were reported in the SUNRISE-1 extension, though post-marketing reports have since appeared in the FDA adverse event reporting system.

Regression to Mean: How Much of the "Improvement" Was Real?

Every insomnia trial faces this criticism. Patients enroll during a period of poor sleep (they must meet severity criteria). Some improvement is expected simply because sleep quality fluctuates and enrollment captures the trough.

The SUNRISE-1 core trial addressed this partially through its placebo arm. Placebo-subtracted LPS reduction was 7-10 minutes, confirming a true drug effect beyond regression to mean.

The extension phase lost this control. With no placebo arm, the sustained improvement at month 12 contains an unknown regression-to-mean component. We cannot cleanly separate "drug still works" from "patient's insomnia would have improved anyway."

SUNRISE-2 partially addresses this because it maintained a placebo arm for 6 months. Lemborexant's superiority over placebo persisted through month 6 on both LPS and WASO (wake after sleep onset), suggesting the drug's contribution is durable and not merely captured natural fluctuation.

Limitations the Authors Acknowledged

The extension data have specific weaknesses that constrain interpretation:

Open-label design introduces expectation bias. Patients knew they were receiving active drug.
Survivor bias is inherent. Only patients who tolerated and responded to lemborexant in the core phase entered the extension. The 30% who did not continue likely included non-responders and those with adverse effects.
No active comparator in the extension. We cannot say whether lemborexant maintained efficacy better or worse than chronic zolpidem or eszopiclone over 12 months.
Subjective endpoints only in the extension. PSG was not repeated, so we lack objective confirmation that the diary-reported improvements reflected true polysomnographic changes.
Single extension without re-randomization means we cannot distinguish continued drug effect from persistent placebo response or lifestyle adaptation.

Clinical Translation: What This Means for Prescribers

The combined SUNRISE-1 core and extension data support prescribing lemborexant with the expectation of sustained benefit over at least 12 months. The American Academy of Sleep Medicine 2023 guidelines conditionally recommend orexin receptor antagonists for chronic insomnia, citing this durability evidence as part of the rationale.

For patients already on lemborexant at 3-6 months who report stable benefit, the extension data suggest no pharmacological reason to attempt dose reduction or drug holiday. Discontinuation can be abrupt without expected rebound, though patients should be counseled that 1-2 nights of mildly worsened sleep may occur.

For patients concerned about long-term safety, the 12-month data and subsequent post-marketing surveillance through 2024 have not revealed hepatotoxicity, cognitive decline, cancer signal, or dependency patterns. The orexin pathway remains mechanistically distinct from GABA sedative risks.

Frequently asked questions

›

References

Rosenberg R, Murphy P, Zammit G, et al. Comparison of Lemborexant With Placebo and Zolpidem Tartrate Extended Release for the Treatment of Older Adults With Insomnia Disorder: A Phase 3 Randomized Clinical Trial. JAMA Netw Open. 2019;2(12):e1918254. https://pubmed.ncbi.nlm.nih.gov/31886325/
Kärppä M, Yardley J, Pinner K, et al. Long-term efficacy and tolerability of lemborexant compared with placebo in adults with insomnia disorder: results from the phase 3 randomized clinical trial SUNRISE 2. Sleep. 2020;43(9):zsaa123. https://pubmed.ncbi.nlm.nih.gov/32025989/
Dayvigo (lemborexant) prescribing information. Eisai Inc. Revised 2022. https://www.accessdata.fda.gov/drugsatfda_docs/label/2022/212028s004lbl.pdf
Sateia MJ, Buysse DJ, Krystal AD, et al. Clinical Practice Guideline for the Pharmacologic Treatment of Chronic Insomnia in Adults: An American Academy of Sleep Medicine Clinical Practice Guideline. J Clin Sleep Med. 2017;13(2):307-349. https://pubmed.ncbi.nlm.nih.gov/27998379/
Herring WJ, Connor KM, Ivgy-May N, et al. Suvorexant in Patients With Insomnia: Results From Two 3-Month Randomized Controlled Clinical Trials. Biol Psychiatry. 2016;79(2):136-148. https://pubmed.ncbi.nlm.nih.gov/25117004/
Mysliwiec V, Martin JL, Engles J, et al. The Management of Chronic Insomnia Disorder and Obstructive Sleep Apnea: Synopsis of the 2023 Updated Practice Parameters. J Clin Sleep Med. 2023;19(4):769-779. https://pubmed.ncbi.nlm.nih.gov/36722697/