Assess the effect of lemborexant (LEM), a dual orexin receptor antagonist, on objective polysomnography-derived sleep measures in elderly subjects with insomnia disorder.

SUNRISE-1 (NCT02783729; E2006-G000-304) was a randomized, double-blind, placebo- and active-controlled, 1-month, global phase 3 study. Females aged ≥55y and males aged ≥65y with insomnia disorder (chief complaint: sleep maintenance difficulties) participated. This analysis evaluated LEM efficacy and safety in the subgroup of subjects aged ≥65y.

After an ~2-week placebo run-in, subjects (n=1006) were randomized to placebo, zolpidem tartrate extended release (ZOL, 6.25mg) or LEM (5mg [LEM5]; 10mg [LEM10]) for 1 month. Averages from paired polysomnograms at baseline (during run-in), Nights 1/2, and Nights 29/30 were analyzed for latency to persistent sleep (LPS), wake after sleep onset (WASO), WASO in the second half of the night (WASO2H), and sleep efficiency (SE; total sleep time/time in bed, standardized at 8h). All subjects met criteria for sleep maintenance difficulties.

453 subjects were aged ≥65y (137 males, 316 females); 95.6% (433/453) completed the study. Significantly greater decreases from baseline (improvement) in LPS, WASO, and WASO2H were observed with LEM5 and LEM10 versus placebo and ZOL at Nights 1/2 (all P<0.05). At Nights 29/30, significantly greater decreases were observed in LPS with LEM versus placebo (LEM10 only, P<0.05) and versus ZOL (both P<0.05), and with both LEM doses versus placebo and ZOL for WASO and WASO2H (all P<0.05). Additionally, increases (improvement) in SE were greater with LEM5 and LEM10 versus placebo and ZOL at both time points (all P<0.05). Most treatment-emergent adverse events (TEAEs) were mild/moderate. No deaths or serious TEAES occurred in these subjects.

In subjects aged ≥65y, LEM provided significant benefit versus placebo and ZOL on sleep onset and sleep maintenance outcomes, and was well tolerated, supporting the development of LEM as a treatment for insomnia disorder in the elderly.