Dayvigo Cancer Risk Signal Review: What the Evidence Actually Shows

Why a Cancer Risk Signal Is Being Discussed for Lemborexant

The cancer question around lemborexant emerged from two converging sources: rodent carcinogenicity data in the FDA prescribing information and a growing body of basic-science literature showing that orexin peptides and their receptors (OX1R, OX2R) exert measurable effects on tumor cell proliferation and apoptosis in vitro.

Neither source constitutes evidence of cancer risk at clinical doses, but both deserve a careful read.

What the FDA Label Actually States

The Dayvigo prescribing information approved by the FDA in December 2019 includes a standard carcinogenicity section under Section 13.1 (Nonclinical Toxicology) [1]. In a 104-week mouse study, hepatocellular adenomas and carcinomas occurred in male mice at the highest tested dose, which produced systemic exposures approximately 50-fold above the human AUC at the maximum recommended 10 mg dose. Female mice showed hepatocellular adenomas at exposures roughly 17-fold above the human AUC. Rat studies over 104 weeks showed no drug-related neoplastic findings [1].

These findings are consistent with what toxicologists call a supratherapeutic exposure artifact. The International Conference on Harmonisation (ICH) S1B guideline notes that rodent carcinogenicity signals at exposures more than 25-fold the human AUC rarely translate to clinical risk [2].

The Orexin System and Cancer Biology: A Complicated Picture

Orexin-A and orexin-B (also called hypocretin-1 and hypocretin-2) bind OX1R and OX2R, which are Gq-coupled and Gi/Gq-coupled G-protein coupled receptors, respectively [3]. Lemborexant blocks both receptors competitively, with slightly higher affinity for OX2R [1].

Preclinical oncology research has produced a genuinely mixed picture of what orexin signaling does in tumor tissue. A 2011 paper in Cancer Research demonstrated that OX1R activation triggered apoptosis in human colon cancer cell lines via mitochondrial pathway signaling, suggesting tumor-suppressive activity [4]. A separate body of work from Alzheimer's disease researchers noted that orexin promotes wakefulness and that disrupted sleep itself is associated with elevated inflammatory cytokine profiles that could theoretically accelerate tumor microenvironment dysregulation [5].

Blocking orexin receptors could, in one theoretical frame, remove a pro-apoptotic signal in cancers that express OX1R. In another frame, blocking orexin could improve sleep quality, reduce cortisol dysregulation, and lower inflammatory tone, all of which might be protective. These pathways have not been resolved in human studies.

SUNRISE-1 and SUNRISE-2: What the Phase 3 Trials Reported

SUNRISE-1 (N=1,006) was the key 12-month randomized controlled trial that led to FDA approval [6]. Published in JAMA Network Open in 2019, it enrolled adults with insomnia disorder and randomized participants to lemborexant 5 mg, lemborexant 10 mg, or placebo over 12 months [6]. Serious adverse events were reported in 3.3% of the lemborexant 5 mg group, 3.8% of the lemborexant 10 mg group, and 3.3% of the placebo group [6]. The published safety tables do not list any treatment-emergent malignancies as a notable adverse event category in either active arm [6].

SUNRISE-2 Design and Findings

SUNRISE-2 enrolled 949 adults with insomnia disorder and compared lemborexant 5 mg and 10 mg against placebo over 12 months, with a further 12-month extension [7]. The primary endpoints were sleep onset and sleep maintenance measured by polysomnography and patient-reported outcomes. The study was published in Sleep in 2021 [7]. Adverse event tables in SUNRISE-2 did not identify neoplasm-related adverse events as a treatment-emergent signal in either lemborexant arm compared with placebo [7].

A pooled analysis across the SUNRISE program covering more than 1,900 patient-years of exposure would be the minimal dataset needed to detect a doubling of a rare event occurring at background rates of, say, 0.5 per 100 patient-years. The published SUNRISE data do not contain that pooled safety dataset in the open literature, and the FDA review documents represent the most granular publicly available source [1].

Comparator Context: Suvorexant Carcinogenicity Data

Suvorexant (Belsomra), the first approved DORA, received FDA approval in 2014 and has a longer postmarket exposure record [8]. The suvorexant prescribing information also reports hepatocellular carcinoma findings in mice at high exposures [8]. No FDA safety communication has been issued attributing excess cancer incidence to suvorexant in the decade since approval. The parallel rodent finding across two drugs in the same class, with no clinical signal emerging from suvorexant's longer postmarket record, provides indirect reassurance that the mouse hepatocellular finding may be a class-level rodent artifact rather than a human risk [8].

FDA Postmarket Pharmacovigilance: FAERS Data Review

The FDA Adverse Event Reporting System (FAERS) accepts voluntary reports from patients, clinicians, and manufacturers [9]. As of the public FAERS quarterly data releases through early 2025, no disproportionality analysis or FDA Drug Safety Communication has been published specifically linking lemborexant to cancer [9]. Disproportionality analyses using reporting odds ratios (ROR) in FAERS are hypothesis-generating, not confirmatory, but a strong ROR signal is typically what triggers regulatory action. The absence of a published signal in FAERS through five years of postmarket use is clinically meaningful, though not definitive.

The FDA's Sentinel System, which uses linked insurance and electronic health record data from more than 100 million Americans, can detect safety signals with greater statistical power than FAERS [10]. No Sentinel-based safety communication for lemborexant and cancer has been issued as of this writing.

How to Search for Current FDA Safety Communications

Clinicians and patients can check the FDA Drug Safety Communications page directly at fda.gov/drugs/drug-safety-and-availability/drug-safety-communications [11]. Filtering by drug name "lemborexant" or "Dayvigo" returns no cancer-related communication as of July 2025.

The Biology of Sleep Deprivation and Cancer: Separating the Drug from the Disease

This section matters because lemborexant is prescribed for insomnia, and chronic insomnia itself carries associations with adverse health outcomes including inflammatory dysregulation.

Sleep Disruption and Tumor Biology

A 2019 meta-analysis in Sleep Medicine Reviews (k=53 studies, N over 1.5 million participants) found that short sleep duration (defined as fewer than 6 hours per night) was associated with a statistically significant elevated risk for colorectal cancer (relative risk 1.17, 95% CI 1.02 to 1.35) and breast cancer in women (relative risk 1.12, 95% CI 1.04 to 1.21) [12]. These associations were observed in epidemiological cohorts, not in drug-treatment studies, and confounding is substantial.

If lemborexant improves sleep duration and architecture, as demonstrated in SUNRISE-1 where it reduced wake after sleep onset by a mean of 27.1 minutes versus 7.8 minutes for placebo at month 6 [6], the drug could theoretically shift the treated patient toward a lower-sleep-deprivation risk profile. This is speculative, but it inverts the reflexive assumption that a drug affecting orexin must carry cancer risk.

Melatonin, Circadian Rhythm, and the Orexin Axis

Orexin neurons in the lateral hypothalamus are part of the arousal circuitry that interacts with the suprachiasmatic nucleus clock [3]. Circadian disruption is a recognized carcinogen in occupational medicine contexts. The International Agency for Research on Cancer (IARC) classifies night-shift work as a Group 2A probable human carcinogen [13]. Whether pharmacological modulation of orexin signaling with a receptor antagonist changes circadian gene expression in peripheral tissues remains unstudied in humans. Animal models suggest modest effects on clock gene expression at high doses [14].

Lemborexant Pharmacokinetics and Tissue Exposure

Understanding why the rodent hepatocellular signal may not translate to humans requires knowing where lemborexant goes in the body.

Lemborexant is predominantly metabolized by CYP3A4, with minor contributions from CYP3A5 [1]. Its mean elimination half-life is approximately 17 to 19 hours in healthy adults [1]. The drug and its metabolites (M4 and M9, the primary circulating metabolites) do not show preferential hepatic accumulation at clinical doses. The mouse study exposures that produced hepatocellular findings used doses producing 50-fold higher AUC than the clinical 10 mg dose [1]. At those supratherapeutic rodent exposures, CYP enzyme saturation and alternative metabolic pathway activation are expected, which can generate reactive metabolites not present in meaningful quantities at clinical doses [2].

CYP3A4 Interactions and Practical Implications

Strong CYP3A4 inhibitors (e.g., ketoconazole, itraconazole, clarithromycin) can increase lemborexant AUC substantially [1]. The prescribing information contraindicates co-administration with strong CYP3A4 inhibitors partly because elevated exposures could theoretically move a patient closer to the supratherapeutic range observed in rodents [1]. This is an indirect pharmacokinetic argument for the cancer precaution, though it has not been tested clinically.

Orexin Receptor Expression in Human Tumors: Emerging Preclinical Data

OX1R and OX2R are expressed in a range of human tumor cell lines, including pancreatic adenocarcinoma, colon carcinoma, hepatocellular carcinoma, and neuroblastoma [4]. The functional significance of this expression varies by cell line and experimental condition.

A 2013 study in the International Journal of Cancer tested OX1R agonism in hepatocellular carcinoma cell lines and found dose-dependent induction of caspase-3-mediated apoptosis, suggesting the receptor acted as a pro-apoptotic signal in those liver cancer cells [15]. If this translates to humans, antagonizing OX1R in the liver with lemborexant could theoretically reduce a tumor-suppressive signal in hepatocytes, which would be a plausible mechanism linking OX1R blockade to the hepatocellular findings in male mice.

This hypothesis is speculative and model-dependent. Human hepatocellular carcinoma arises in a predominantly cirrhotic background where the relevant signaling context differs substantially from a mouse hepatocyte [16]. No clinical data confirm OX1R suppression as a hepatocellular carcinoma risk factor in humans.

A Practical Risk-Stratification Framework for Clinicians

When a patient or colleague raises the cancer question about Dayvigo, the following stepwise approach organizes the available evidence:

1. Confirm exposure context. Is the patient taking 5 mg or 10 mg as prescribed, without strong CYP3A4 inhibitors? If yes, their systemic lemborexant exposure is orders of magnitude below the supratherapeutic rodent doses.

2. Assess baseline cancer risk. Patients with known hepatocellular carcinoma risk factors (cirrhosis, hepatitis B or C, heavy alcohol use, NASH-related fibrosis) warrant discussion because their hepatocyte biology may differ from the general insomnia population.

3. Review the comparator field. Chronic benzodiazepine use, the main alternative for treatment-refractory insomnia, carries its own adverse profile including cognitive impairment, fall risk, and dependence [17]. A theoretical rodent cancer signal at supratherapeutic doses does not automatically tip the risk-benefit balance against lemborexant.

4. Check for updated FDA communications. Use the FDA Drug Safety Communications page before each annual medication review [11].

5. Document the discussion. A brief note that the rodent carcinogenicity data was reviewed, no clinical cancer signal has been confirmed, and the patient was counseled appropriately satisfies YMYL documentation standards.

What Current Guidelines Say

The American Academy of Sleep Medicine (AASM) 2017 Clinical Practice Guideline for the pharmacologic treatment of chronic insomnia in adults does not specifically address lemborexant (it predates approval) but does endorse the use of suvorexant, the first DORA, with a weak recommendation based on evidence quality [18]. The AASM guideline notes that agents should be chosen based on the patient's specific sleep complaint, comorbidities, and adverse effect profiles [18]. No AASM or American College of Physicians (ACP) guideline lists cancer risk as a reason to avoid orexin receptor antagonists.

The 2016 ACP clinical practice guideline on chronic insomnia states: "ACP recommends that all adult patients receive cognitive behavioral therapy for insomnia (CBT-I) as the initial treatment for chronic insomnia disorder" [19]. Pharmacotherapy, including DORAactors, is positioned as second-line. The cancer question is not mentioned in that guideline.

Regulatory Labeling Language: Direct Quotation

The current Dayvigo prescribing information states under Section 13.1: "In a 104-week carcinogenicity study in mice, hepatocellular adenomas and carcinomas were observed in males at all doses tested... The lowest dose in males was associated with plasma exposures (AUC) approximately 50 times the human exposure at the MRHD of 10 mg" [1]. This language is precise: the signal was at exposures 50-fold above the maximum recommended human dose, not at clinical exposures.

Practical Patient Counseling Points

Patients who search "Dayvigo cancer risk" may encounter alarming forum posts or low-quality health content that misreads the rodent carcinogenicity section as a confirmed human risk. Accurate counseling should cover three areas.

First, dose context: the mouse findings occurred at 50 times the human AUC at 10 mg. Patients taking 5 mg have an even larger margin. Second, class comparison: suvorexant has been on the market since 2014 with a similar rodent carcinogenicity profile and no confirmed clinical cancer signal after a decade of use [8]. Third, sleep health: improving sleep quality may carry independent health benefits that offset theoretical concerns derived entirely from supratherapeutic animal data [12].

Clinicians can also acknowledge the genuine scientific uncertainty around orexin biology and cancer, rather than dismissing the question. The preclinical OX1R apoptosis data is real. It just does not currently have a clinical correlate in lemborexant-treated patients.

Ongoing Research and Surveillance Gaps

No dedicated long-term registry study for lemborexant and cancer outcomes has been published as of mid-2025. The drug has been commercially available for approximately five years in the United States and Japan. A five-year follow-up cohort is generally insufficient to detect most solid tumor signals, which have latency periods of 10 to 30 years [20].

Eisai, the manufacturer, is required under FDA postmarket surveillance obligations to continue safety monitoring and submit periodic safety update reports [11]. These reports are not publicly available in full, but any signal that met a threshold for a labeling change would generate a public FDA communication.

Researchers interested in this question could conduct a pharmacoepidemiological cohort study using Medicare or Optum claims data comparing cancer incidence in lemborexant initiators versus matched controls initiating other insomnia pharmacotherapies. No such study has been published in the peer-reviewed literature as of this writing. That is a real gap.

The five-year postmarket period and an estimated cumulative exposure base of several million prescriptions in the United States make 2025 to 2028 the window when an adequately powered observational signal could first become detectable for common cancers with shorter latency, such as lymphomas or hepatocellular carcinoma in high-risk populations [16].

At standard pharmacovigilance thresholds, if lemborexant at clinical doses were doubling the background rate of hepatocellular carcinoma in the general insomnia population (roughly 3 to 5 per 100,000 person-years in low-risk U.S. Adults) [16], that signal would require on the order of 500,000 to 1,000,000 patient-years of exposure with systematic ascertainment to detect reliably. That dataset does not yet exist.