A 12-Week, Open-Label, Multicenter Pilot Study to Evaluate the Efficacy and Safety of Lemborexant in Patients with Insomnia Comorbid with Depressive Episodes (SELENADE)

Abstract

Background and Objectives

The efficacy and safety of lemborexant, a dual orexin receptor antagonist, for treating insomnia associated with depressive episodes remain unclear. This pilot study aimed to evaluate the changes in clinical symptoms following the initiation of lemborexant treatment and its safety in patients with insomnia comorbid with depressive episodes.

Methods

The inclusion criteria for this multicenter, prospective, interventional, open-label pilot study conducted in Japan were adults (≥ 18 years) diagnosed with insomnia and either major depressive disorder (MDD) or bipolar disorder (BD) according to DSM-5, receiving treatment for a depressive episode for ≥ 4 weeks, and having a baseline Hamilton Depression Rating Scale (HAM-D-17) score of ≥ 8. Patients were classified into four cohorts on the basis of their MDD or BD diagnosis and concomitant insomnia medication use (lemborexant add-on or monotherapy). After a 2-week lead-in period with unchanged prior medications for insomnia MDD or BD, 12-week lemborexant treatment was administered. The primary endpoint was the change from baseline in the Insomnia Severity Index (ISI) total score at week 4, which was analyzed using paired t-tests with a two-sided significance level of 5%. For other endpoints, summary statistics and 95% confidence intervals (CIs) for observed values and changes from baseline were calculated by cohort. Safety outcomes included the evaluation of the safety and tolerability of lemborexant during the treatment period.

Results

A total of 83 patients with comorbid insomnia and depressive episodes were enrolled (MDD add-on cohort, n = 29; MDD monotherapy, n = 23; BD add-on, n = 16; and BD monotherapy, n = 15). The average age was 47.2 years and 64.6% were female. Mean baseline ISI total score (SD) was 14.5 (4.8), 13.2 (5.5), 13.5 (6.0), and 11.8 (4.6), respectively. At week 4, mean change from baseline in the ISI total score [95% CI, p-value] was − 2.3 [− 3.9, − 0.7, p = 0.0065], − 3.2 [− 5.0, − 1.3, p = 0.0018], − 5.2 [− 7.9, − 2.5, p = 0.0012], − 4.5 [− 6.3, − 2.6, p = 0.0001], respectively. Improvements were sustained through week 12. Sleep diary measures such as sleep onset latency and total sleep time improved in some cohorts, but the results were inconsistent. Mean baseline HAM-D-17 scores were 13.6 (5.1), 12.6 (5.1), 13.9 (4.3), and 11.4 (3.4), respectively, with mean changes at week 12 of − 6.4 [− 7.8, − 4.9], − 6.2 [− 7.7, − 4.6], − 5.2 [− 8.4, − 2.1], and − 5.4 [− 7.2, − 3.5], respectively. No serious treatment-emergent adverse events (TEAEs) were reported, with TEAE incidence ranging from 30% to 40%, all mild to moderate in severity. Somnolence was the most common adverse reaction, reported in 10.3–25.0% of cohorts except BD monotherapy. Nightmare followed, occurring in 4.3–12.5% across cohorts.

Conclusions

Lemborexant was associated with improvements in insomnia severity and depressive symptoms related to MDD or BD, with no new safety concerns. It may be a treatment option for residual insomnia following treatment for depression associated with MDD or BD, though somnolence and nightmares warrant attention. Confirmatory studies are needed to establish its therapeutic value.

Study Registration

Japan registry of clinical trials (jRCTs071220037; registered on 4 August 2022)

Supplementary Information

The online version contains supplementary material available at 10.1007/s40263-025-01245-w.

Key Points

In patients with insomnia comorbid with depressive episodes, lemborexant was associated with modest improvements in insomnia symptoms from week 4 of treatment; its effects were generally sustained through week 12.

Depressive symptoms associated with major depressive disorder or bipolar disorder also improved after the initiation of lemborexant treatment.

There were no major safety concerns identified with lemborexant.

Introduction

Insomnia is a disorder characterized by symptoms such as difficulty falling asleep, difficulty maintaining sleep, and early morning awakening, all of which cause distress and impair daytime functioning. Insomnia is commonly comorbid with mood disorders, primarily major depressive disorder (MDD) and bipolar disorder (BD) [1, 2]. Insomnia was previously considered as one of the symptoms of depressive episodes; however, it often persists even after depressive episodes have remitted [3] and is increasingly recognized as an independent disorder [4, 5]. Despite this, insomnia and MDD or BD are closely interconnected, with insomnia influencing the prognosis of treatment for MDD or BD [6, 7]. In particular, residual insomnia following mood disorder treatment is associated with a higher risk of relapse [3], highlighting the importance of effective sleep management during remission, which is considered essential in the treatment of MDD and BD.

Pharmacologic treatments for insomnia include benzodiazepine (BZ) receptor agonists, a class that includes BZs and non-BZ drugs (Z-drugs), melatonin receptor agonists, and orexin receptor antagonists. Although Z-drugs have demonstrated efficacy [8–11], safety concerns persist, particularly regarding the risk of tolerance, dependence, abuse, and withdrawal. This is especially relevant for patients with mood-disorder-associated insomnia. However, the evidence regarding the efficacy of pharmacologic treatments for insomnia comorbid with BD remains limited.

Recent evidence suggests that dysregulation of the orexin system may contribute to the pathophysiology of MDD [12, 13]. Orexins, neuropeptides involved in sleep–wake regulation, arousal, and emotional processing act via two receptors (OX1R and OX2R) widely distributed in the brain. While preclinical findings regarding the antidepressant effect of orexin antagonism remain inconsistent, clinical trials suggest that selective antagonism of OX2R may offer more consistent antidepressant benefits than OX1R antagonism [14]. For example, seltorexant, a selective OX2R antagonist, has shown antidepressant effects independent of sleep improvement in patients with MDD [15, 16]. Furthermore, naturalistic studies of daridorexant, a dual orexin receptor antagonist (DORA), suggest that targeting insomnia may improve not only sleep, but also mood symptoms, emotion regulation, and executive function in patients with comorbid psychiatric conditions [17, 18]. These data support the hypothesis that modulation of orexin signaling may offer a novel approach to treating depression, particularly in patients with coexisting insomnia.

Lemborexant is a DORA that competitively antagonizes orexin binding at OX1R and OX2R. In two phase 3 clinical trials that involved patients with insomnia (Study E2006-G000-303 [Study 303; NCT02952820] [19] and Study E2006-G000-304 [Study 304; NCT02783729] [20]), lemborexant demonstrated efficacy on sleep onset and sleep maintenance variables, as well as exhibited a well-tolerated safety profile. However, patients with moderate-to-severe depressive symptoms were excluded from these trials, limiting the investigation of lemborexant’s effects on insomnia associated with depressive episodes. Therefore, obtaining data on lemborexant’s efficacy and safety in patients with insomnia comorbid with depressive episodes of MDD and BD holds significant scientific importance.

In this study, we evaluated the changes in insomnia severity, depressive and manic symptoms, subjective sleep parameters during 12 weeks of lemborexant treatment, participants’ global impression of insomnia medications, and the safety of lemborexant in Japan in patients with insomnia comorbid with depressive episodes of MDD or BD.

Methods

Study Design

This was a multicenter, prospective, interventional, open-label pilot study to evaluate the efficacy and safety of lemborexant in patients with insomnia comorbid with depressive episodes conducted in Japan. The study began enrolling participants in August 2022 and continued until the completion of observation for the last study participant in June 2024. The study consisted of two phases: a 2-week pretreatment period (the baseline period) and a 12-week treatment period during which lemborexant was administered. During the pretreatment period, participants underwent eligibility screening and baseline assessments. Those who met the study criteria proceeded to the treatment period and received lemborexant for 12 weeks.

Study Population

The main inclusion criteria were as follows: male and female participants aged 18 years or older at the time of consent; those who met the diagnostic criteria for insomnia disorder as defined in the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5) [21]; patients meeting the DSM-5 diagnostic criteria for a depressive episode and a diagnosis of either MDD or BD, while receiving stable treatment for a depressive episode for at least 4 weeks prior to the pretreatment period; and patients having a total score of 8 or higher on the 17-item Hamilton Rating Scale for Depression (HAM-D-17) at baseline.

The main exclusion criteria were as follows: moderate or severe obstructive sleep apnea; periodic limb movement disorder, restless legs syndrome, or parasomnias; comorbidity or history of narcolepsy or cataplexy; or experience of a manic or hypomanic episode within 4 weeks before the start of the pretreatment period.

Eligible patients were recruited from existing outpatient populations at each site on the basis of information from electronic medical records. Study participants were assigned to one of four cohorts on the basis of comorbid mood disorders and the presence or absence of prior medications for insomnia. The MDD and BD add-on cohorts included participants diagnosed with MDD or BD, respectively, who had been using one insomnia medication other than suvorexant (the only DORA available in Japan at the time of the study) before the pretreatment period began. The MDD and BD monotherapy cohorts included participants who were diagnosed with MDD or BD, respectively, and who had no prior medication for insomnia.

Treatment

Lemborexant was initiated at 5 mg/day and maintained for 4 weeks. The dose could be adjusted up to a maximum of 10 mg/day at the discretion of the investigators [22]. The initiation of new or modification of treatment for mood disorders and the initiation of any new treatment for insomnia were prohibited throughout the study period. In the add-on cohort, the same dose of the previous insomnia medication was continued for the first 4 weeks of the treatment period. Thereafter, a reduction in the dose of the prior medication was recommended.

Endpoints and Assessments

The primary endpoint was the change in insomnia severity from baseline, as measured by the total score of the Insomnia Severity Index (ISI) [23–25] at week 4 of lemborexant treatment. Secondary endpoints included changes from baseline in ISI at weeks 8 and 12, and in mood disorder severity, subjective sleep parameters, and patients’ global impression of treatment effects at weeks 4, 8, and 12.

The ISI is a 7-item self-administered questionnaire that assesses the type and severity of insomnia symptoms over the past 2 weeks. Patients rate each item on a 5-point Likert scale, with higher scores indicating greater severity. The total score ranges from 0 to 28. The ISI was assessed at baseline and at weeks 4, 8, and 12.

The HAM-D-17 [26, 27] assesses the severity of depression over the past week, with higher scores indicating more severe depression. The total score ranges from 0 to 52 and includes three items that evaluate insomnia symptoms (items 4–6). A total score excluding sleep-related items (range 0–46) was also calculated. The HAM-D-17 was administered by trained raters at baseline and at weeks 4, 8, and 12.

The Young Mania Rating Scale (YMRS) [28] is an 11-item questionnaire used to assess the severity of manic episodes, with total scores ranging from 0 to 60. Higher scores indicate greater severity. The YMRS was administered by trained raters at baseline and at weeks 4, 8, and 12 in patients in the BD add-on and BD monotherapy cohorts. This assessment was conducted to detect manic switching.

The Patient Global Impression–Insomnia version (PGI-I) [29, 30] is a self-administered questionnaire designed to assess the overall impression of treatment effects on sleep. Participants responded to three items regarding treatment effects (“treatment helped me sleep,” “treatment reduced time to fall asleep,” and “treatment increased total sleep time”) using “positive,” “negative,” or “neutral.” One additional item assessed the appropriateness of perceived treatment strength (“too strong,” “just right,” or “too weak”). The PGI-I was assessed at baseline and at weeks 4, 8, and 12. Participants evaluated their prescribed insomnia medication as a whole. At baseline, those in the add-on cohort assessed their previous treatment. From that point onward, participants in the monotherapy cohort evaluated lemborexant alone, while those in the add-on cohort evaluated the combination of lemborexant and their prior insomnia medication.

Subjective sleep parameters were recorded using a sleep diary during the pretreatment period and the first 4 weeks of treatment. Participants recorded sleep parameters and sleep quality from the previous night, as well as morning alertness and mood, immediately after waking. Subjective nighttime sleep parameters included subjective sleep onset latency (sSOL), subjective wake after sleep onset (sWASO), subjective number of awakenings, subjective total sleep time (sTST), and subjective sleep efficiency (sSE). sSOL was defined as the time from attempting to fall asleep to sleep onset; sWASO as the total wake time after sleep onset until leaving bed, including early morning awakenings; and sSE as the proportion of time in bed spent asleep (i.e., sTST divided by time in bed). Sleep quality, next-morning sleepiness/alertness, and mood were rated on a 9-point Likert scale (1, worst; 9, best).

Safety Assessment

Secondary endpoints also included safety outcomes, specifically the incidence, severity, and nature of treatment-emergent adverse events (TEAEs) during the 12-week lemborexant treatment period. TEAEs were assessed at each visit and categorized by severity (mild, moderate, or severe) and judged for causal relationship to lemborexant by the investigators. Adverse events were coded using Medical Dictionary for Regulatory Activities (MedDRA) version 27.0 preferred terms.

Statistical Analysis

We used SAS (Ver. 9.4) to conduct analyses after the database was locked, in accordance with a statistical analysis plan that had been finalized prior to database lock. The full analysis set (FAS) was defined on the basis of the intent-to-treat principle and included all study participants who received at least one dose of lemborexant and had at least one postdose efficacy evaluation. Participants without follow-up assessments were excluded due to the inability to evaluate outcomes. We defined safety analysis set as all study participants who received at least one dose of lemborexant.

Efficacy Analysis

The efficacy analyses were conducted using the FAS. For continuous variables, we presented the mean and standard deviation (SD) of the observed values, the mean and SD of the change from baseline, and the 95% confidence interval (CI) for each cohort. For categorical variables, we calculated the number of participants, percentages, and 95% CIs for each cohort. The CIs for the mean values were estimated using the t-distribution, whereas those for proportions were estimated using the Clopper–Pearson method (exact method). We analyzed the primary endpoint using a paired t-test to compare baseline and week 4 values with a two-sided significance level of 5%. Sleep diary variables were calculated on the basis of entries recorded during the 7 days preceding baseline and week 4. Data were considered missing if fewer than 3 days of entries were available.

A multivariate logistic regression analysis was performed to explore independent predictors of treatment response. Treatment response was defined as a reduction of seven points or more in the ISI total score from baseline at week 12. Patients who did not meet this criterion or discontinued the study were classified as nonresponders. The dependent variable was responder status, and the following independent variables were included: comorbid mood disorder (MDD versus BD), lemborexant treatment modality (add-on versus monotherapy), sex, age, duration of insomnia, duration of the comorbid mood disorder, baseline ISI total score, and baseline HAM-D-17 total score. Both forced-entry and stepwise selection methods were applied. For the stepwise procedure, the significance level for entry and retention in the model was set at 0.15.

Safety Analysis

We conducted the safety analysis using the safety analysis set. The number of participants and percentages for each TEAE were reported by cohort.

Results

Participants

A total of 87 patients provided consent at 10 medical institutions across Japan. However, 4 patients withdrew prior to enrollment, resulting in the enrollment of 83 participants: 29 in the MDD add-on cohort, 23 in the MDD monotherapy cohort, 16 in the BD add-on cohort, and 15 in the BD monotherapy cohort. All participants initiated protocol treatment and were included in the safety analysis set. Of these, a total of 74 participants completed the study: 25 in the MDD add-on cohort, 20 in the MDD monotherapy cohort, 14 in the BD add-on cohort, and 15 in the BD monotherapy cohort. One patient from the BD add-on cohort who had been using ramelteon was excluded from the FAS, resulting in a total of 82 participants (Fig. 1).

Fig. 1.

Patient disposition flow chart. The primary reasons for study discontinuation were listed. MDD major depressive disorder, BD bipolar disorder

The proportion of female participants in each cohort ranged from 53.3% to 73.9%, with ages between 43.7 and 51.5 years (Table 1). Baseline mean ± SD ISI total scores were 14.5 ± 4.8 for the MDD add-on cohort, 13.2 ± 5.5 for the MDD monotherapy cohort, 13.5 ± 6.0 for the BD add-on cohort, and 11.8 ± 4.6 for the BD monotherapy cohort (Supplementary Table 1 in Online Resource 1). Baseline HAM-D-17 total scores ranged from 8 to 28, with mean ± SD of 13.6 ± 5.1 for the MDD add-on cohort, 12.6 ± 5.1 for the MDD monotherapy cohort, 13.9 ± 4.3 for the BD add-on cohort, and 11.4 ± 3.4 for the BD monotherapy cohort (Supplementary Table 2 in Online Resource 1). The mean baseline YMRS total scores ± SD were 1.1 ± 1.8 for the BD add-on cohort and 2.3 ± 2.5 for the BD monotherapy cohort (Supplementary Table 3 in Online Resource 1).

Table 1.

Baseline demographics and characteristics (full analysis set)

		MDD add-on cohort	MDD monotherapy cohort	BD add-on cohort	BD monotherapy cohort
		n = 29	n = 23	n = 15	n = 15
Sex, female	n (%)	17 (58.6)	17 (73.9)	8 (53.3)	11 (73.3)
Age (years)	Mean (SD)	45.0 (13.8)	49.6 (17.2)	51.5 (13.5)	43.7 (17.1)
Age group, ≥ 65 years	n (%)	1 (3.4)	3 (13.0)	2 (13.3)	1 (6.7)
Age at onset of insomnia (years)	Mean (SD)	39.5 (12.0)	41.4 (16.9)	34.6 (13.9)	31.9 (16.8)
Duration of insomnia (years)a	Mean (SD)	5.6 (5.6)	8.2 (10.8)	15.4 (14.0)	10.4 (6.5)
Age at onset of mood disorder (years)	Mean (SD)	37.0 (13.6)	42.5 (17.2)	34.1 (12.2)	31.3 (15.9)
Duration of mood disorder (years)b	Mean (SD)	8.0 (6.5)	7.3 (11.0)	17.4 (14.3)	11.1 (5.7)
Bipolar disorder (BD-I)	n (%)	–	–	5 (33.3)	4 (26.7)
Other comorbid psychiatric disorders
Autism spectrum disorder	n (%)	2 (6.9)	0 (0.0)	1 (6.7)	1 (6.7)
Attention-deficit hyperactivity disorder	n (%)	0 (0.0)	1 (4.3)	1 (6.7)	0 (0.0)
Work status
Full-time workc	n (%)	5 (17.2)	4 (17.4)	2 (13.3)	4 (26.7)
Shorter hours, part-time workd	n (%)	4 (13.8)	0 (0.0)	3 (20.0)	4 (26.7)
Leave of absence/unemployed	n (%)	20 (69.0)	18 (78.3)	10 (66.7)	6 (40.0)
Experience of absence from work, yes	n (%)	22 (75.9)	14 (60.9)	8 (53.3)	7 (46.7)

SD standard deviation, MDD major depressive disorder,BD bipolar disorder

^aDuration of insomnia (years) = [age (years)] − [age at onset of insomnia (years)]

^bDuration of mood disorder (years) = [age (years)] − [age at onset of mood disorder (years)]

^cFull-time work is defined as work at least 8 h per day, at least 4 days per week, or equivalent paid work

^dShorter hours, part-time work is defined as part-time work, regular paid work of at least 1 day per week, including commuting to a workshop.

Concomitant medications for psychiatric disorders, including insomnia and mood disorders, are summarized in Supplementary Table 4 (Online Resource 1). All participants were receiving treatment for comorbid MDD or BD. The most frequently used medications included mirtazapine (range 7.5–45 mg), trazodone (25–150 mg), escitalopram (10 or 20 mg), lurasidone (20–60 mg), lithium carbonate (200–1200 mg), and vortioxetine (10 or 20 mg). Two participants underwent changes in medications for mood disorders during the study period: one in the MDD monotherapy cohort had sertraline titrated from 50 mg to 100 mg between visit 2 and visit 3; the other in the MDD add-on cohort had venlafaxine reduced from 187.5 mg to 150 mg at visit 4. All other concomitant medications in the FAS were administered in accordance with the protocol. The most frequently used concomitant insomnia medications in the add-on cohorts were eszopiclone (1–3 mg), brotizolam (0.125–0.5 mg), and flunitrazepam (0.25–2 mg) (Supplementary Table 4 in Online Resource 1). No participants received cognitive behavioral therapy for insomnia.

Efficacy

At week 12, the percentage of participants continuing lemborexant was 100% (29/29) for MDD add-on cohort, 73.9% (17/23) for MDD monotherapy cohort, 100% (16/16) for BD add-on cohort, and 86.7% (13/15) for BD monotherapy cohort. The proportion of participants who adhered to lemborexant on more than 80% of study days was 100% (29/29), 73.9% (17/23), 100% (16/16), and 86.7% (13/15), respectively, and the mean ± SD daily dose of lemborexant was 5.78 ± 1.18 mg/day, 5.34 ± 1.68 mg/day, 5.64 ± 1.27 mg/day, and 5.12 ± 0.78 mg/day, respectively (Supplementary Table 5 in Online Resource 1). The percentage of participants who required dose increase of lemborexant was 37.9% (11/29), 39.1% (9/23), 57.1% (8/14), and 13.3% (2/15), respectively.

The mean changes ± SD in ISI total score from baseline at week 4 were statistically significant across all cohorts: − 2.3 ± 4.2 (95% CI − 3.9 to − 0.7, p = 0.0065) for the MDD add-on cohort, − 3.2 ± 4.2 (95% CI − 5.0 to − 1.3, p = 0.0018) for the MDD monotherapy cohort, − 5.2 ± 4.7 (95% CI − 7.9 to − 2.5, p = 0.0012) for the BD add-on cohort, and − 4.5 ± 3.4 (95% CI − 6.3 to − 2.6, p = 0.0001) for the BD monotherapy cohort. At weeks 8 and 12, continued reductions in ISI scores were observed in all cohorts, with 95% CIs remaining below 0, except for week 8 in the BD add-on cohort (mean ± SD, − 3.7 ± 6.8; 95% CI − 7.8 to 0.4) (Fig. 2, Supplementary Table 1 in Online Resource 1). This trend was consistent regardless of whether lemborexant dose escalation was required (Supplementary Table 6 in Online Resource 1). At week 12, the percentage of responders, defined as individuals with a reduction in ISI score from baseline of seven points or more, was 41.4% (12/29) in the MDD add-on cohort, 19% (4/21) in the MDD monotherapy cohort, 35.7% (5/14) in the BD add-on cohort, and 42.9% (6/15) in the BD monotherapy cohort. To identify factors associated with responder status at week 12, a stepwise multivariate logistic regression analysis was conducted (Table 2). The final model retained four variables: bipolar disorder diagnosis (versus MDD), older age, shorter duration of mood disorder, and higher baseline ISI score. The respective odds ratios (95% confidence intervals) were 3.43 (0.95–12.42), 1.03 (0.99–1.07), 0.90 (0.83–0.98), and 1.22 (1.07–1.38).

Fig. 2.

Changes from baseline in the Insomnia Severity Index (ISI) score (7 items) after lemborexant treatment. The primary endpoint was the change in ISI score from baseline to week 4. Paired t-test showed p < 0.05 for every cohort. No statistical tests were conducted at other time points. Error bars indicate 95% confidential intervals of means. ISI insomnia severity index, MDD major depressive disorder, BD bipolar disorder, CI confidential interval, BL baseline, W4, W8, or W12 4 weeks, 8 weeks, or 12 weeks after administration of the study drug, respectively

Table 2.

Multivariate logistic regression analysis

Factor		N	Responder	%	Force entry method Adjusted odds ratio (95% CI)	Stepwise method Adjusted odds ratio (95% CI)
Disease
MDD		48	15	31.3	(Reference)	(Reference)
BD		26	10	38.5	3.40 (0.86–13.36)	3.43 (0.95–12.42)
Lemborexant treatment
Monotherapy		33	10	30.3	(Reference)	(Reference)
Add-on		41	15	36.6	1.36 (0.43–4.33)	–
Sex
Male		24	10	41.7	(Reference)	(Reference)
Female		50	15	30.0	0.49 (0.14–1.69)	–
Age (years)		74	25	33.8	1.03 (0.99–1.07)	1.03 (0.99–1.07)
Duration of insomnia (years)		74	25	33.8	1.08 (0.94–1.24)	–
Duration of mood disorder (years)		74	25	33.8	0.85 (0.74–0.97)	0.90 (0.83–0.98)
ISI total score at baseline		74	25	33.8	1.20 (1.05–1.37)	1.22 (1.07–1.38)
HAM-D-17 total score at baseline		74	25	33.8	1.03 (0.91–1.17)	–

MDD major depressive disorder, BD bipolar disorder, ISI insomnia severity index, HAM-D-17 Hamilton Depression Rating Scale 17 items, CI confidential interval

The mean HAM-D-17 total score decreased from baseline, with 95% CI remaining below 0 at all time points in each cohort. At week 12, the mean change ± SD from baseline in the HAM-D-17 total score was − 6.4 ± 3.5 (95% CI − 7.8 to − 4.9) in the MDD add-on cohort, − 6.2 ± 3.1 (95% CI − 7.7 to − 4.6) in the MDD monotherapy cohort, − 5.2 ± 5.2 (95% CI − 8.4 to − 2.1) in the BD add-on cohort, and − 5.4 ± 3.2 (95% CI − 7.2 to − 3.5) in the BD monotherapy cohort. We observed a similar trend in the HAM-D-17 total score excluding the sleep-related items (Fig. 3, Supplementary Table 2 in Online Resource 1).

Fig. 3.

Changes from baseline for the HAM-D-17 score (17 items) (a) and the HAM-D-17 subtotal score (excluding sleep-related items) (b). Error bars indicate 95% confidential intervals of means. HAM-D-17 Hamilton Rating Scale for Depression, 17 items, MDD major depressive disorder, BD bipolar disorder, CI confidential interval, BL baseline, W4, W8, or W12 4 weeks, 8 weeks, or 12 weeks after administration of the study drug, respectively

No participants had a YMRS score of four at any assessment timepoint, indicating that all participants remained in complete remission of manic state [31] after the initiation of lemborexant treatment (Supplementary Table 3 in Online Resource 1).

Positive responses to the PGI-I items “treatment helped me sleep,” “treatment reduced time to fall asleep,” and “treatment increased total sleep time” increased from baseline at week 4 in the add-on cohorts, and this trend persisted through week 12. Even among the monotherapy cohorts, most responses remained positive throughout the lemborexant treatment period. The proportion of patients who rated the appropriateness of treatment intensity as “just right” exceeded 60% at all timepoints in all cohorts (Supplementary Fig. 1 in Online Resource 1).

Among nighttime insomnia symptoms, including sSOL, sWASO, subjective number of awakenings, sTST, and subjective sleep efficiency, sSOL showed a consistent trend toward improvement: mean ± SD, − 24.6 ± 44.9 min (95% CI − 43.1 to − 6.1) in the MDD add-on cohort, − 16.5 ± 30.6 min (95% CI − 30.9 to − 2.2) in the MDD monotherapy cohort, − 13.3 ± 34.4 min (95% CI − 33.2 to 6.5) in the BD add-on cohort, and − 4.9 ± 20.3 min (95% CI − 17.1 to 7.3) in the BD monotherapy cohort. In contrast, sWASO showed no consistent trend toward improvement: mean ± SD, − 16.9 ± 47.2 min (95% CI − 2.6 to 36.4), − 0.4 ± 48.7 min (95% CI − 23.1 to 23.9), − 37.9 ± 60.2 min (95% CI − 74.3 to − 1.5), and − 7.2 ± 58.4 (95% CI − 28.1 to 42.5) in the respective cohorts (see Supplementary Table 7 in Online Resource 1). There was an numerical increase in morning alertness and quality of sleep: mean ± SD, 0.3 ± 1.1 (−95% CI 0.1 to 0.8), 0.5 ± 0.9 (95% CI 0.1 to 1.0), 0.0 ± 1.1 (95% CI − 0.6 to 0.6), and 0.2 ± 1.4 (95% CI − 0.7 to 1.0), respectively, for morning alertness; and 0.4 ± 1.2 (95% CI − 0.1 to 0.9), 0.6 ± 0.9 (95% CI 0.1 to 1.0), 0.3 ± 0.7 (95% CI 0.0 to 0.7), and 0.5 ± 1.6 (95% CI − 0.5 to 1.4), respectively, for quality of sleep.

Safety

The incidence of TEAEs was 41.4% (12/29) in the MDD add-on cohort, 43.5% (10/23) in the MDD monotherapy cohort, 37.5% (6/16) in the BD add-on cohort, and 33.3% (5/15) in the BD monotherapy cohort. We did not identify any serious adverse events in this study, and the severity of all TEAEs was mild to moderate. Most frequent TEAEs that led to lemborexant discontinuation were reported in three participants in the MDD add-on cohort. Of these TEAEs, the investigator considered abdominal discomfort and rash to be causally related to lemborexant, whereas dyspepsia was deemed unrelated. Somnolence was the most frequently reported adverse reaction to lemborexant, occurring in 10.3% (3/29) of the MDD add-on cohort, 17.4% (4/23) of the MDD monotherapy cohort, 25.0% (4/16) of the BD add-on cohort, and 0.0% (0/15) of the BD monotherapy cohort. Nightmare was the second most common adverse reaction, occurring in 6.9% (2/29) of the MDD add-on cohort, 4.3% (1/23) of the MDD monotherapy cohort, 12.5% (2/16) of the BD add-on cohort, and 6.7% (1/15) of the BD monotherapy cohort (Table 3).

Table 3.

Treatment-emergent adverse events (TEAEs) and adverse drug reactions

	MDD add-on cohort		MDD monotherapy cohort		BD add-on cohort		BD monotherapy cohort
	n = 29		n = 23		n = 16		n = 15
	n	(%)	n	(%)	n	(%)	n	(%)
TEAEs	12	(41.4)	10	(43.5)	6	(37.5)	5	(33.3)
Mild TEAEs	11	(37.9)	7	(30.4)	6	(37.5)	4	(26.7)
Moderate TEAEs	3	(10.3)	3	(13.0)	1	(6.3)	1	(6.7)
TEAEs reported by > 1 participant
Nasopharyngitis	2	(6.9)	1	(4.3)	0	(0.0)	1	(6.7)
Nightmare	2	(6.9)	1	(4.3)	2	(12.5)	1	(6.7)
Dizziness	0	(0.0)	2	(8.7)	0	(0.0)	0	(0.0)
Somnolence	4	(13.8)	4	(17.4)	4	(25.0)	0	(0.0)
Malaise	1	(3.4)	1	(4.3)	0	(0.0)	0	(0.0)
TEAEs leading to the discontinuation of the study drug	3	(10.3)	0	(0.0)	0	(0.0)	0	(0.0)
Abdominal discomfort	1	(3.4)	0	(0.0)	0	(0.0)	0	(0.0)
Dyspepsia	1	(3.4)	0	(0.0)	0	(0.0)	0	(0.0)
Rash	1	(3.4)	0	(0.0)	0	(0.0)	0	(0.0)
TEAEs leading to death	0	(0.0)	0	(0.0)	0	(0.0)	0	(0.0)
Serious AE observed after administration of the study drug	0	(0.0)	0	(0.0)	0	(0.0)	0	(0.0)
Adverse drug reactions (reported by > 1 participant)	8	(27.6)	7	(30.4)	6	(37.5)	2	(13.3)
Nightmare	2	(6.9)	1	(4.3)	2	(12.5)	1	(6.7)
Dizziness	0	(0.0)	2	(8.7)	0	(0.0)	0	(0.0)
Somnolence	3	(10.3)	4	(17.4)	4	(25.0)	0	(0.0)
Malaise	1	(3.4)	1	(4.3)	0	(0.0)	0	(0.0)

TEAEs and adverse drug reactions reported in at least two patients in total are listed

MDD major depressive disorder, BD bipolar disorder, TEAE treatment-emergent adverse event, AE adverse event

Discussion

To the best of our knowledge, this was the first prospective study conducted to investigate the efficacy and safety of lemborexant in participants with insomnia comorbid with MDD- and BD-related depressive symptoms [32]. Treatment with lemborexant was associated with improvements not only in insomnia, but also in depressive symptoms related to MDD and BD, and was well tolerated. While the absence of a control group limits definitive conclusions regarding efficacy, the study provides exploratory insights into the use of lemborexant in a real-world clinical population.

Participants exhibited baseline insomnia symptoms generally classified as mild (8–14 points) on the basis of the ISI total score. The average severity of depressive symptoms was also mild (8–16 points) based on the HAM-D-17 total score [33], although patients with more severe depression (HAM-D-17 up to 28) were also included. We considered the study population representative of patients with comorbid insomnia who, despite receiving stable treatment for mood disorders, had not achieved remission of depressive symptoms associated with MDD or BD. Given that residual insomnia following mood disorder treatment has been associated with increased risk of depressive relapse and poorer long-term outcomes [3, 6, 7], the inclusion of such patients in this study highlights a clinically relevant target population.

Insomnia symptoms showed statistically significant improvement at week 4 of the lemborexant treatment; these effects were generally maintained through week 12. Although the change from baseline was modest and did not reach the clinically meaningful threshold (≥ 7-point reduction), the mean score at week 12 ranged from 6.0 to 8.9 across the groups, approaching the normal range of ≤ 7 points [34]. Logistic regression analysis identified higher baseline insomnia severity as a factor associated with treatment response. Generally, greater symptom severity at baseline allows more room for improvement, potentially resulting in larger treatment effects [35, 36]. Since the participants in this study had relatively mild symptoms, this may have contributed to the modest changes observed. The logistic regression analysis also showed that shorter duration of comorbid mood disorders was associated with better treatment response. Mood disorders such as MDD and BD have been reported to become less responsive to treatment as the duration of illness increases [37, 38]. In the present study, a similar trend was observed in the pharmacological treatment of insomnia among patients with MDD or BD. Although concomitant treatments for mood disorders were stable before entering the study, all participants were receiving such treatments, which could represent a potential confounding factor. Nevertheless, the majority of assessment measures suggested that participants perceived an improvement in their sleep.

Depressive symptoms associated with MDD or BD also showed a statistically significant improvement from baseline [39]. The HAM-D-17 total score gradually decreased toward week 12, with changes at week 12 (−5.2 to − 6.4) aligning with thresholds for clinically meaningful improvement [39]. In the MDD add-on, MDD monotherapy, and BD monotherapy cohorts, the score fell below the 7-point remission threshold [33] at week 12. Notably, improvement in HAM-D-17 scores was observed even when sleep-related items were excluded, suggesting that lemborexant may have potential antidepressant effects. This observation aligns with emerging evidence that orexin receptor antagonists may modulate mood-related pathways. For example, seltorexant, a selective orexin-2 antagonist, demonstrated antidepressant efficacy compared with placebo in randomized controlled trials (N = 47) [16]. Although lemborexant is primarily indicated for insomnia, its potential impact on depressive symptoms warrants further investigation in controlled studies.

As we excluded patients with BD with manic symptoms from this study, the mean baseline YMRS total score remained less than 12, the clinical remission threshold, in all participants during the 12 weeks of lemborexant treatment, indicating that lemborexant did not affect manic symptoms.

Compared with previous studies [8, 17, 18] investigating patients with comorbid mood disorders and insomnia, the participants in this study exhibited milder depressive symptoms associated with MDD or BD. In a previous randomized controlled study, which included 545 patients with HAM-D-17 scores of 22.1–22.4 on average, the combination of fluoxetine and eszopiclone significantly reduced the HAM-D-17 total scores, excluding sleep-related items, as compared with fluoxetine plus placebo (changes from baseline at week 8: eszopiclone, − 12.9; placebo, − 10.9 for all items; eszopiclone, − 9.5; placebo, − 8.4 for scores excluding sleep-related items). In contrast, our study population had baseline HAM-D-17 scores ranging from 11.4 to 13.9, and the observed improvements at week 12 ranged from − 5.2 to − 6.4. As with the ISI findings discussed above, the relatively mild baseline severity may have contributed to the smaller magnitude of change, which nonetheless remained within the range of clinically meaningful improvement. Recent naturalistic studies using daridorexant, another DORA, have also reported improvements from baseline in insomnia severity and depressive symptoms in patients with comorbid mood disorders including MDD and BD (N = 66 and 90) [17, 18]. These findings suggest that DORAs may exert broader therapeutic effects beyond sleep regulation, particularly in patients with mild-to-moderate depressive symptoms. Taken together, the observed improvements in depressive symptoms—despite the absence of targeted antidepressant intervention—support the need for further investigation into the role of sleep-focused treatments in mood disorder populations.

Although the overall insomnia severity, as measured by the ISI, showed improvement by week 4 of the lemborexant treatment, the subjective sleep parameters did not consistently reflect this effect. This discrepancy might be attributed to the variability in insomnia symptoms (e.g., difficulty falling asleep and waking after sleep onset) among the study participants, as the inclusion/exclusion criteria did not account for subjective sleep parameters. Nonetheless, sSOL showed consistent improvement trend across cohorts. This suggests that, in the present study population, the improvement in insomnia severity may have been primarily driven by the alleviation of sleep initiation difficulties. Although previous studies have shown variability in the associations between specific insomnia symptoms and depressive outcomes or long-term prognosis [40, 41], some have identified sSOL as most strongly linked to the onset of depression [42–44]. The present findings suggest that the rapid onset of action of lemborexant [45] may offer particular benefit in treatments targeting sleep onset difficulties.

In the present study, 33.3–43.5% of participants experienced adverse events, with somnolence and nightmares being the most frequently reported. These findings are consistent with a previous 3-month prospective study of lemborexant conducted in Japan [46]. Post-marketing surveillance data have shown that the risk of somnolence is higher in patients receiving concomitant medications [47]. All participants in this study were taking at least one additional psychotropic medication, which may have contributed to the increased incidence of somnolence. Additionally, patients with depression have been reported to experience nightmares more frequently than those without depressive disorders [48]. To ensure the safe use of lemborexant, concomitant medications and comorbid conditions should be carefully considered. No serious TEAEs were observed in this study, and all reported TEAEs were mild or moderate in severity. These findings indicate that lemborexant does not pose a major safety concern as a treatment for insomnia comorbid with mood disorders.

This study’s strength is that it is the first prospective intervention study of a DORA for insomnia comorbid with BD and MDD. Few intervention studies have examined therapeutic agents for insomnia comorbid with BD, and our study offers a valuable contribution to the field. This study also has several limitations that should be considered when interpreting the findings. First, this employed an open-label design without control group. This limits the ability to draw causal inferences regarding the efficacy of lemborexant. Second, as a pilot study, the sample size of 15–30 participants per cohort was determined on the basis of site feasibility rather than statistical power. Consequently, sufficient statistical evaluation of the endpoints could not be conducted. Third, all participants received various concomitant medications. Although the protocol restricted changes in mood disorder treatments, some sedative agents, such as trazodone, may have influenced sleep outcomes. Furthermore, two patients underwent changes during the treatment period, which could have confounded the results. Fourth, the assessment of insomnia symptoms relied primarily on subjective measures, such as the ISI and sleep diary. These instruments are susceptible to recall bias and expectancy effects, which may have led to overestimation of treatment effects. Fifth, most participants in this study had mild depressive symptoms associated with MDD or BD. The generalizability of the findings to individuals with moderate-to-severe depression remains limited. Future randomized controlled trials with larger sample size in broader target patients are warranted to confirm the efficacy and safety of lemborexant. Sixth, although we included several potential confounding factors such as baseline insomnia severity and duration of mood disorder in our regression analysis, other relevant variables, such as the number of previous depressive episodes, total duration of pharmacological treatment, and history of psychotherapy, were not collected in this study. Finally, this study was conducted exclusively in Japan, which may limit the generalizability of the findings to other populations.

Conclusions

Lemborexant treatment was associated with improvements in both insomnia severity and depressive symptoms related to MDD or BD. No new safety concerns were identified. These findings suggest that lemborexant may be considered as a treatment option for patients commonly encountered in clinical practice who experience residual insomnia following treatment for depression. In such cases, particular attention should be paid to the potential emergence of somnolence and nightmares. Further confirmatory studies are warranted to establish its therapeutic value.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

This study was funded by Eisai Co., Ltd. The sponsor participated in the study design, including the analysis plan and the interpretation of the results. The sponsor provided funding to EPS Corporation for the conduct of the study, monitoring, data collection and management, statistical analysis, auditing, and manuscript writing support. All authors, including those affiliated with the sponsor, met the authorship criteria, participated in the review and approval of the manuscript, and contributed to the decision to submit the manuscript for publication. This open access fee for this publication was funded by Eisai Co., Ltd.

Declarations

Conflict of Interest

Y.Taka. has received lecture fees from Eisai, Takeda Pharmaceutical, Sumitomo Pharma, Otsuka Pharmaceutical, Mochida Pharmaceutical, Lundbeck Japan, Viatris Pharmaceuticals Japan, Nobelpharma, Meiji Seika Pharma, MSD, Daiichi Sankyo, and Shionogi & Co. outside the submitted work. H.M. has received honoraria from Eisai, Otsuka Pharmaceutical, Mochida Pharmaceutical, Sumitomo Pharma, Takeda Pharmaceutical, Kyowa Kirin, MSD, Meiji Seika Pharma, Janssen Pharmaceutical, Yoshitomi Pharmaceutical Industries, and Viatris Pharmaceuticals Japan. M.T. has received honoraria from Takeda Pharmaceutical, Otsuka Pharmaceutical, Daiichi Sankyo, Sumitomo Pharma, Meiji Seika Pharma, Viatris Pharmaceuticals, MSD, Eisai, and Yoshitomi Pharmaceutical. M.Ka. has received consulting fees from Sumitomo Pharma, Otsuka Pharmaceutical, Lundbeck Japan, and Shionogi & Co. and honoraria from Sumitomo Pharma, Eisai, Janssen Pharmaceutical, Meiji Seika Pharma, Otsuka Pharmaceutical, Takeda Pharmaceutical, Viatris, Mitsubishi Tanabe Pharma, MSD, Eli Lilly, Shionogi & Co., Lundbeck Japan, Kyowa Pharmaceutical, and Ono Pharmaceutical. H.Hi. received all support for the present manuscript from Eisai. H.Ho. received all support for the present manuscript from Eisai, and received honoraria from Sumitomo Pharma, Eisai, Janssen Pharmaceutical, Meiji Seika Pharma, Otsuka Pharmaceutical, Takeda Pharmaceutical, and Viatris. K.I. received all support for the present manuscript from Eisai, and honoraria from Eisai, MSD, Otsuka Pharmaceutical, Kyowa Kirin, Kowa Pharmaceutical, Shionogi & Co., Sumitomo Pharma, Daiichi Sankyo, Takeda Pharmaceutical, Mitsubishi Tanabe Pharma, Eli Lilly Japan, Boehringer Ingelheim Japan, Novartis Pharma, Pfizer Japan, Meiji Seika Pharma, Mochida Pharmaceutical, Janssen Pharmaceutical, Lundbeck Japan, Yoshitomi Pharmaceutical Industries, Ono Pharmaceutical, Nipro, Nexera Pharma, EA Pharma, and Viatris Pharmaceuticals Japan. H.S. has received grants from Japan Society for the Promotion of Science, Japan Research Foundation Clinical Pharmacology, and Takeda Science Foundation, and honorarium from Viatris, Eisai, Takeda Pharmaceutical, Otsuka Pharmaceutical, Meiji Seika Pharma, Shionogi Pharma, Yoshitomi Pharmaceutical Industries, Sumitomo Pharma, Kyowa Pharmaceutical, MSD, Mochida Pharmaceutical, and Lundbeck Japan. M.O. has received grants from Mitsubishi Tanabe Pharma, Tsumura, Shionogi & Co., Mochida Pharmaceutical, Sumitomo Pharma, Otsuka Pharmaceutical, Kyowa Kirin, Teijin Pharma and honoraria from Eisai, Tsumura, MSD, Meiji Seika Pharma, Takeda Pharmaceutical, Nobel pharma, Taisho Pharmaceutical, Daiichi Sankyo, and Shionogi & Co. Y.Ko. received all support for the present manuscript from Eisai; grants from JSPS KAKENHI; consulting fees from Takeda Pharmaceutical; and honoraria from Takeda Pharmaceutical, Lundbeck Japan, Sumitomo Pharma, Otsuka Pharmaceutical, and Sanwa Kagaku Kenkyusho. T.M. has received honoraria from Sumitomo Pharma, Otsuka Pharmaceutical, Takeda Pharmaceutical, Viatris, and Shionogi & Co. T.T. has received lecture fees from Sumitomo Pharma, Daiichi Sankyo, Viatris, Eisai, Mochida Pharmaceutical, and Shionogi & Co. K.W. received all support for the present manuscript; received grants from Eisai, Meiji Seika Pharma, Mitsubishi Tanabe Pharma, Mochida Pharmaceutical, MSD, Otsuka Pharmaceutical, Sumitomo Dainippon Pharma, and Takeda Pharmaceutical; consulting fees from Boehringer Ingelheim, Daiichi Sankyo, Eisai, Janssen Pharmaceutical, Kyowa Pharmaceutical, Lundbeck Japan, Luye Pharma, Mitsubishi Tanabe Pharma, Otsuka Pharmaceutical, Sumitomo Dainippon Pharma, Taisho Toyama Pharmaceutical, Takeda Pharmaceutical, and Viatris; honoraria from Boehringer Ingelheim, Eisai, Janssen Pharmaceutical, Kyowa Pharmaceutical, Lundbeck Japan, Meiji Seika Pharma, Mitsubishi Tanabe Pharma, MSD, Otsuka Pharmaceutical, Shionogi, Sumitomo Dainippon Pharma, Takeda Pharmaceutical, and Viatris; and has participated on DSMB or advisory board of Boehringer Ingelheim, Mitsubishi Tanabe Pharma, Nippon Chemiphar, and Otsuka Pharmaceutical. H.Y. received all support for the present manuscript from Eisai. M.Ko. and Y.Ka. are full-time employees of Eisai Co., Ltd. Y.Take. has received grant funding from the Japan Society for the Promotion of Science and speaker’s honoraria from Eisai, Meiji Seika Pharma, Sumitomo Pharma, Janssen Pharmaceutical, Otsuka, Lundbeck, Daiichi Sankyo, Takeda Pharmaceutical, UCB Japan, Novartis, Teijin Pharma, Nippon Boehringer Ingelheim, and Mitsubishi Tanabe Pharma.

Ethics Approval

The study protocol, protocol amendments, and all study-related documents were reviewed and approved by the review board of the University of the Ryukyus for clinical research (approval no.: 19-2). Study registration number is jRCTs071220037. The study was conducted in accordance with the protocol, and complied with the Clinical Trials Act, the Declaration of Helsinki, and all local laws.

Consent to Participate

All patients provided written informed consent to participate in the study.

Consent to Publication

Not applicable.

Availability of Data And Material

The dataset generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Code Availability

No custom code or scripts were developed specifically for this study.

Author Contributions

Conceptualization and design of the study: Y.Taka., T.T., K.W., and M.Ko. Data collection and investigation: Y.Taka., H.M., M.T., M.Ka., H.Hi., H.Ho., K.I., H.S., M.O., Y.Ko., T.M., T.T., K.W., H.Y., and Y.Take. Project management: Y.Taka., M.Ko., and Y.Ka. Writing and revising of the original draft: Y.Taka., M.Ko., and Y.Ka. Writing review and editing: Y.Taka., H.M., M.T., M.Ka., H.Hi., H.Ho., K.I., H.S., M.O., Y.Ko., T.M., T.T., K.W., H.Y., M.Ko., Y.Ka., and Y.Take. All authors have read and approved the final submitted manuscript and agree to be accountable for the work.