Background: Some sleep-promoting medications are associated with cognitive impairment, making treatment of comorbid obstructive-sleep-apnea (OSA) and insomnia (COMISA) challenging. Lemborexant is a dual-orexin-receptor-antagonist approved for insomnia treatment. This post-hoc analysis evaluated cognition in the subgroup of subjects with mild-OSA (apnea-hypopnea-index ≥5–<15 events/h-of-sleep). Methods: Study E2006-G000-304 was a 1-month, randomized, double-blind, placebo (PBO)- and active-comparator (zolpidem-ER 6.25mg [ZOL])-controlled study of lemborexant 5/10mg (LEM5/LEM10). Subjects ≥55y with insomnia disorder/sleep maintenance problems were enrolled (N=1006). A cognitive-performance assessment battery (CPAB) was performed at morning waketime of Days(D)2/3 and D30/31. Change-from-baseline (CFB) for mean power-of-attention (PoA), continuity-of-attention (CoA), quality-of-memory (QoM), and speed-of-memory-retrieval (SoMR) for CPAB tasks was analyzed. Results: The mild-OSA subgroup comprised 410 (40.8%) subjects. On D2/3 and D30/31, CFB for PoA, CoA, QoM, and SoMR for LEM5/LEM10 were not significantly different than PBO. On D2/3, PoA and QoM were significantly worse with ZOL vs PBO; QoM was significantly better with LEM5/LEM10 vs ZOL. On D30/31, SoMR was significantly worse with ZOL vs PBO and significantly better with LEM5/LEM10 vs ZOL. LEM safety in the subgroup was consistent with the overall study population. Conclusions: Memory and attention domains in subjects with COMISA characterized by mild-OSA were not impacted by LEM, unlike ZOL.

Support: Eisai Inc.