Belsomra (Suvorexant) History and Development: From Orexin Discovery to FDA Approval

The Discovery of Orexins and the Scientific Foundation

The entire scientific premise behind suvorexant traces back to 1998, when two independent research groups identified the orexin neuropeptide system. Luis de Lecea and colleagues described hypocretins in rat hypothalamus, while Takeshi Sakurai and colleagues at the University of Texas Southwestern identified orexin-A and orexin-B as ligands for two orphan G-protein-coupled receptors 1.

These findings arrived at almost the same time. Within months, the connection between orexin deficiency and narcolepsy became clear. In 1999, Emmanuel Mignot's group at Stanford demonstrated that canine narcolepsy resulted from mutations in the orexin-2 receptor gene 2. Simultaneously, Masashi Yanagisawa's team showed that orexin-knockout mice exhibited a narcolepsy-like phenotype with sudden behavioral arrests 3.

The logic was straightforward. If orexin deficiency causes excessive sleepiness and narcolepsy, then blocking orexin signaling pharmacologically could promote sleep in patients with insomnia. This "reverse narcolepsy" hypothesis gave drug developers a biologically rational target for the first time in insomnia pharmacology, moving beyond the broad CNS depression produced by benzodiazepines and GABAergic Z-drugs 4.

Merck's Preclinical Development Program

Merck's insomnia program began screening orexin receptor antagonists in the early 2000s. The medicinal chemistry effort was substantial. Researchers needed a compound that crossed the blood-brain barrier efficiently, bound both OX1R and OX2R with high affinity, and had a half-life suitable for a full night of sleep without next-morning hangover.

Early Merck publications described diazepane-based scaffolds optimized through structure-activity relationship studies 5. The team evaluated hundreds of analogs. Suvorexant (MK-4305) emerged as the lead clinical candidate by approximately 2007, selected for its balanced receptor binding (Ki of 0.55 nM at OX1R and 0.35 nM at OX2R), oral bioavailability, and appropriate pharmacokinetic profile 6.

Preclinical studies in rats and dogs confirmed that suvorexant reduced active-wake time and increased both REM and non-REM sleep in a dose-dependent manner. Unlike benzodiazepines, the compound did not suppress REM sleep. This was a notable pharmacologic distinction. Animal models also showed that suvorexant's sleep-promoting effects were reversible: when presented with a salient stimulus, animals could be aroused, suggesting the drug reduced wake drive rather than imposing sedation 6.

Phase 1 and Phase 2 Clinical Trials

Suvorexant entered human testing around 2007. The phase 1 program established safety, tolerability, and basic pharmacokinetics. The drug demonstrated a terminal half-life of approximately 12 hours, with peak plasma concentrations reached at about 2 hours post-dose 7.

The first published proof-of-concept trial was a randomized, double-blind, placebo-controlled crossover study in 254 patients with primary insomnia, reported by Herring and colleagues in 2012 7. Patients received suvorexant at doses of 10 mg, 20 mg, 40 mg, or 80 mg for four weeks. The results were significant. At night 1, the 40 mg and 80 mg doses improved polysomnography-measured sleep efficiency by 8.4 and 10.3 percentage points versus placebo, respectively. All doses from 20 mg upward improved subjective total sleep time at four weeks.

These phase 2 data confirmed the orexin antagonist hypothesis in humans. They also introduced a dose-response puzzle that would complicate the drug's regulatory path: higher doses showed greater efficacy, but safety reviewers later raised concerns about next-day residual effects at those higher exposures.

The Key Phase 3 Trials

Merck's registration program comprised two large, multicenter, randomized, double-blind, placebo-controlled phase 3 trials, designated Study 028 and Study 029. Published together by Herring et al. in The Lancet Neurology in 2014, these trials enrolled 3,291 patients with insomnia aged 18 and older 8.

Both trials used polysomnography (PSG) endpoints in the first month and patient-reported outcomes over three months. The tested doses were 40 mg for non-elderly adults and 30 mg for elderly patients (age 65 and older), with a low-dose comparator arm of 20 mg and 15 mg, respectively.

The primary endpoints were time to sleep onset (latency to persistent sleep, or LPS) and wake after sleep onset (WASO) measured by PSG. At month 1, the higher-dose groups showed a mean reduction in WASO of 22.7 minutes in Study 028 and 16.1 minutes in Study 029 compared to placebo. Sleep onset latency improved by 7.6 minutes and 8.4 minutes in the two studies, respectively 8.

Patient-reported total sleep time (sTST) also improved. By month 3, patients on the higher dose reported sleeping 20 to 25 minutes longer per night than those on placebo. The lower-dose arms (20 mg/15 mg) showed consistent, though smaller, effects on all endpoints. No rebound insomnia was observed after discontinuation, and there was no evidence of withdrawal symptoms upon stopping the drug.

Adverse events were generally mild. The most common was somnolence, occurring in approximately 7% of patients on the higher dose versus 3% on placebo. No cases of cataplexy were reported, addressing a theoretical concern given the drug's mechanism. Sleep paralysis occurred in <1% of treated patients.

The FDA Review: A Contentious Approval

Merck filed a New Drug Application (NDA) for suvorexant in August 2013. The review process became one of the more closely watched regulatory episodes in sleep medicine. The original NDA proposed doses of 40 mg for adults and 30 mg for elderly patients, the same doses studied in the phase 3 trials.

The FDA's advisory committee convened on May 22, 2013 (prior to formal NDA submission, as part of the end-of-phase-2 interaction framework). Reviewers expressed concern about next-morning impairment at the 40 mg and 30 mg doses, particularly regarding driving ability. FDA pharmacology reviewers noted that suvorexant's 12-hour half-life meant significant plasma levels persisted into morning hours at higher doses 9.

Dr. Russell Katz, then director of the FDA's Division of Neurology Products, commented during the advisory committee meeting that the agency needed to "balance the sleep benefits at bedtime against the potential for residual next-day effects." The committee voted 13-3 to recommend approval, but with a strong recommendation for lower starting doses.

The FDA approved suvorexant on August 13, 2014, but at significantly lower doses than Merck had tested in phase 3: the approved range was 5 mg to 20 mg, with a recommended starting dose of 10 mg 9. The 30 mg and 40 mg doses were not approved. This represented an unusual regulatory outcome where the approved dosing was below what the key trials had primarily studied. The DEA subsequently placed suvorexant into Schedule IV.

Dr. Andrew Krystal, then at Duke University and a prominent clinical trialist in sleep medicine, noted in a subsequent review: "The approval of suvorexant validated the orexin antagonist mechanism but left the field with a drug whose approved doses were lower than those generating the most compelling efficacy data" 10.

How Suvorexant Works: The Dual Orexin Receptor Antagonist Mechanism

Suvorexant blocks both orexin receptor subtypes (OX1R and OX2R) expressed in wake-promoting brain regions, including the locus coeruleus, tuberomammillary nucleus, dorsal raphe, and lateral hypothalamus. By occupying these receptors, suvorexant prevents orexin-A and orexin-B from activating the downstream monoaminergic, cholinergic, and histaminergic circuits that sustain wakefulness 4.

This mechanism differs from older insomnia medications in a fundamental way. Benzodiazepines and Z-drugs (zolpidem, eszopiclone, zaleplon) enhance GABAergic inhibition globally across the brain. They force sleep by amplifying the brain's primary inhibitory system. Orexin antagonists, by contrast, reduce wake drive by removing excitatory input to arousal centers. The distinction matters clinically. GABAergic drugs tend to suppress slow-wave and REM sleep architecture, while orexin antagonists preserve physiological sleep staging 11.

A 2014 systematic review and meta-analysis by Kishi et al. confirmed that suvorexant improved sleep onset and maintenance measures versus placebo with a favorable side-effect profile, reporting no significant increase in falls, cognitive impairment, or complex sleep behaviors at approved doses 12.

The receptor pharmacology also explains the drug's safety profile regarding abuse potential. A human abuse liability study showed that suvorexant at 40 mg, 80 mg, and 150 mg produced "drug liking" scores below those of zolpidem 30 mg, though supratherapeutic doses did show some subjective effects 13.

Post-Approval Data and Real-World Experience

Following approval, several studies expanded the evidence base for suvorexant at its approved doses. A 2016 analysis by Herring et al. pooled data from 554 patients treated with suvorexant 15 mg or 20 mg and demonstrated sustained improvements in sTST and WASO over 12 months without evidence of tolerance 14.

Post-marketing pharmacovigilance data supported the drug's safety. The FDA label was updated with warnings about next-morning impairment and sleep-related complex behaviors (consistent with class-wide labeling for all sedative-hypnotics), but no unique safety signals emerged beyond what the clinical trial program had identified.

Real-world prescription data showed a more measured adoption than might be expected for a first-in-class agent. Multiple factors contributed: the approved dose ceiling limited efficacy compared to phase 3 results, the Schedule IV designation imposed prescribing friction, and the branded price (approximately $350 to $400 per month without insurance at launch) constrained access. Generic suvorexant became available in the United States after patent litigation was resolved, broadening access.

A Japanese post-marketing surveillance study enrolling 3,234 patients confirmed that suvorexant at 15 mg and 20 mg improved Insomnia Severity Index scores from a mean baseline of 17.2 to 10.8 at 12 weeks, with a discontinuation rate of 9.3% due to adverse events 15.

Suvorexant's Legacy: Opening the DORA Class

Suvorexant's approval in 2014 proved the orexin antagonist concept was viable. Two subsequent DORAs reached the market. Lemborexant (Dayvigo, Eisai) received FDA approval in December 2019, and it was studied at doses specifically chosen to optimize the efficacy-to-residual-effect ratio that had complicated suvorexant's review 16. The SUNRISE-2 trial (N=949) showed lemborexant 5 mg and 10 mg improved sleep onset and maintenance over 12 months versus placebo.

Merck itself developed a second DORA. Suvorexant's patent estate and the lessons learned from its regulatory path informed the development of these follow-on compounds. The class now appears in the American Academy of Sleep Medicine (AASM) 2023 practice guidelines, which conditionally recommend orexin receptor antagonists for sleep-onset and sleep-maintenance insomnia in adults 17.

Dr. Emmanuel Mignot, whose narcolepsy genetics work laid the scientific groundwork for the entire class, reflected that "the development of orexin antagonists for insomnia is a textbook example of bedside-to-bench-to-bedside translational medicine, starting with the clinical observation of narcolepsy, moving through receptor biology, and returning to patients with a rationally designed therapeutic" 10.

The recommended dose of suvorexant remains 10 mg, increased to 20 mg if needed, taken no more than once per night within 30 minutes of bedtime, with at least 7 hours remaining before planned awakening 9.