68-1

16/98

14signaling is necessary for normal circadian regula-tion of consolidated wakefulness. Accordingly, as compared to melatonin, orexin is predominantly secreted during daytime in humans13). In contrast to melatonin, it has been reported that orexin secretion decreases by 10% only in older adults as compared to younger adults15).Interestingly, as compared to controls, patients with moderate to severe Alzheimer disease were shown to have higher mean orexin levels in CSF. Additionally, these patients had significantly impaired nocturnal sleep compared to controls and patients with mild Alzheimer’s disease16). Further-more, significantly higher levels of orexin in the brain of rats with acute pancreatitis were report-edly found when compared to healthy controls17). As dementia and inflammation are risk factors for delirium, it is possible that patients with delirium have increased orexin levels, thus accounting for the resultant sleep-wake disturbance18).Suvorexant, a potent and highly selective for orexin-1 receptor and orexin-2 receptor antagonist receptor antagonist, is an FDA approved treatment for primary insomnia in U.S. Suvorexant has been associated with improvements in subjective measures of total sleep time (sTST), time to sleep onset (sTSO), and wake after sleep onset (sWASO)19), without altering NREM and REM sleep architec-ture as assessed by electroencephalographic moni-toring20). The scientific rationale for using suvorexant for delirium prevention was that such character of suvorexant promoting natural sleep could improve sleep-wake cycle disturbance in delirium. High selec-tivity for orexin-1 and orexin-2 receptor antago-nism21) and little affinity for acetylcholine receptors (Ki>10µM)22) may be advantage for delirium prevention. Reportedly, another dual orexin receptor antagonist did not lower hippocampal acetylcho-line23). Thus, we hypothesized that suvorexant would have effects of preventing delirium.To assess the efficacy of suvorexant in preven-tion of delirium, our group conducted a multi-center, rater-blinded, placebo-controlled RCT in ICUs and acute-phase wards. Eligible patients were aged 65-89 years, who were newly admitted due to medical and surgical emergency, able to take medicine orally, and expected to stay less than 48 hours. Study participants were randomly assigned to suvorexant (15 mg/day) (n=36) or placebo (n=36) nightly for 3 days. Main outcome measure was incidence of delirium according to the DSM-5. Patients taking suvorexant developed delirium less frequently than those taking placebo (suvorexant, 0% (n/N = 0/36) vs. placebo, 17% (6/36), P=0.025)24). With respect to changes in sleep-wake cycle disturbance score (item #1) of DRS-R-98, analysis of variance (ANOVA) revealed a tendency for main effect of treatment, suggesting the potential of suvorexant to improve sleep-wake cycle disturbance that is a core feature of delirium.Another RCT examining the effects suvorexant on delirium prevention in ICU setting showed that, as compared to placebo, suvorexant led to signifi-cant reduction in incidence of both clinical delirium (14.7% vs. 33.3%, P=0.069) and sub-syndromal delirium symptoms (17.6% vs. 47.2%, P=0.011)25). Furthermore, Kaplan-Meier estimates revealed that time to delirium onset was significantly longer in the suvorexant group as compared to the placebo group. These findings, taken together, suggest that suvorexant, a potent and selective orexin antago-nist, has beneficial effects on delirium prevention, and highlight the importance of correcting sleep-wake cycle disturbance in prevention of delirium.Recent meta-analysis shows effectiveness of orexin receptor antagonists for delirium prevention26).4. Real-world effectiveness of melatonin receptor agonists and orexin receptor antagonists After success of RCTs on delirium prevention by ramelteon and suvorexant, we choose these drugs for patients who have risk factors for delirium in clinical practice. So, we examined whether ramel-teon and/or suvorexant would affect delirium prevention among both patients at risk for but without delirium (patients-at-risk), and those with delirium (patients-with-delirium) on the night before a consultation. This multicenter, prospec-tive, observational study was conducted by trained psychiatrists as consultation-liaison psychiatric services between October 2017 and October 2018. Patients who were age 65 years or older and hospi-talized because of acute diseases or elective surgery, had risk factors for delirium, and had insomnia or delirium on the night before the consul-tation were prescribed ramelteon and/or suvorexant. The decision to take medication was left to the for delirium prevention