Lemborexant

Eisai's DORA — the only sleep drug that has objectively beaten zolpidem-ER head-to-head (SUNRISE-1) on sleep onset, with strongest sleep-onset signal in the 2025 DORA network meta-analysis (LEM10 SMD −0.430). But the 17-19 hr half-life is the longest in the DORA class — counter-intuitively the WORST property for a brain-priority late-chronotype 20-year-old who needs total clearance before 8 AM training. For Dylan: SKIP-FOR-NOW. If he ever needs pharmacological sleep escalation, daridorexant (8 hr) clears cleaner; lemborexant is the right pick only if sleep onset is the dominant complaint AND morning grogginess is acceptable.

Lemborexant is a competitive, reversible dual antagonist at both orexin receptors (OX1 and OX2) — same class as daridorexant and suvorexant.

The orexin system in plain English: Orexin (also called hypocretin) is a hypothalamic neuropeptide that functions as the brain's "stay awake" master switch. Orexin neurons live in the lateral hypothalamus, fire vigorously during wakefulness, fall silent during sleep, and project broadly to every wake-promoting nucleus — locus coeruleus (norepinephrine), tuberomammillary nucleus (histamine), raphe (serotonin), VTA (dopamine), basal forebrain (acetylcholine). When orexin neurons fire, they keep all these wake systems online. When they shut up, sleep happens. DORAs occupy the receptors and block orexin's wake-promoting input — so sleep is permitted by removing one major arousal driver, not forced by global GABA-A activation (the Z-drug/benzo mechanism).

Lemborexant-specific pharmacology:

• Half-life ~17-19 hours (mean ~17 hr at 5 mg, ~19 hr at 10 mg). This is the longest in the approved DORA class — suvorexant 12 hr, daridorexant 8 hr. Designed for sleep maintenance through the second half of the night and early-morning awakenings, but at the cost of next-day residual at the high end of the dosing range.
• Tmax ~1-3 hours (median ~1 hr fasted). Onset is rapid — this is the structural reason lemborexant scores best on sleep-onset latency in the NMA despite the long terminal half-life.
• Off-rate at OX1 vs OX2: Eisai's mechanistic paper (Yoshida et al., 2014) reported lemborexant has a faster dissociation rate at OX1 than OX2 — i.e., functionally biased toward sustained OX2 antagonism with shorter OX1 occupancy. This was framed by Eisai as "more OX2-selective in terms of duration" and pitched as mechanistically advantageous because OX2 is the dominant sleep-promoting target. The debate: Some independent re-analyses (Nature Translational Psychiatry 2025 and other reviews) view this as marketing-overstated — at therapeutic concentrations, both receptors are occupied for most of the night, and the clinical efficacy profile (vs daridorexant, suvorexant) doesn't cleanly map to the kinetic difference. Treat the OX2-selectivity claim as plausible but not load-bearing.
• CYP3A4-dominant metabolism. Like daridorexant, primary clearance is hepatic CYP3A4. Multiple inactive and at least one active metabolite (M10) at low circulating concentrations.
• No active metabolite at clinically meaningful levels for the parent pharmacology — the long half-life is the parent compound, not metabolite-driven.

Eisai development context: Eisai (Tokyo) developed E2006/lemborexant after suvorexant (Merck) demonstrated DORA-class viability. Phase 3 program ran 2016-2018; FDA approved December 2019; DEA scheduled IV June 2020. Co-developed in some markets with Purdue Pharma (US co-promotion at launch). Approved Japan (2020), Canada (2020), Australia, Taiwan, Brazil. Not centrally approved by EMA as of 2026-05 — this matters for European sourcing but is irrelevant for Dylan in the US. Eisai has also pushed lemborexant into adjacent indications, most notably Irregular Sleep-Wake Rhythm Disorder (ISWRD) in mild-to-moderate Alzheimer's (E2006-G000-303 / E2006-G000-304 program) — see Evidence below.

The half-life paradox: Lemborexant's mechanistic appeal is paradoxical. The long half-life should mean better sleep-maintenance and worse next-day cognition. The clinical data partially supports this (slight edge on WASO over time, more reports of next-day somnolence than daridorexant), but ALSO shows the strongest sleep-onset effect in the 2025 NMA — which is counter-intuitive for a long-half-life drug. The reconciling explanation is the rapid Tmax: lemborexant gets to therapeutic occupancy fast, then lingers. For Dylan, the relevant question is the lingering, not the onset.