Introduction: Study 303 (SUNRISE-2; NCT02952820) evaluated lemborexant (LEM) efficacy in subjects with insomnia. This post-hoc analysis examined sleep outcomes in the subgroup of subjects with a depression history vs subjects without a depression history (all other subjects).
Methods: Study 303 was a randomized, double-blind, 12-month global phase 3 study. Subjects (≥18y) with insomnia disorder were randomized to placebo or LEM (5mg [LEM5]; 10mg [LEM10]) for 6mo. For the second 6mo, placebo subjects were rerandomized to LEM5 or LEM10; LEM subjects continued at the same dose (reported separately). Subjects with a history of depression, permitted concomitant antidepressant medication use, and/or mild depression (Beck Depression Inventory-II score 14-19) could participate. Least squares mean (LSM) change from baseline (CFB) values were based on a mixed-effect repeated measurement model. For subjective sleep onset latency (sSOL), differences were assessed using least squares geometric means.
Results: Of 949 subjects (placebo, n=318; LEM5, n=316; LEM10, n=315), 112 had a depression history (placebo, n=34; LEM5, n=39; LEM10, n=39). At 6mo, in both subgroups for LEM vs placebo, median CFB (min) in sSOL was greater (depression history: placebo, −12.9; LEM5, −21.7; LEM10, −40.1; all other subjects: placebo, −11.1; LEM5, −21.9; LEM10, −26.4). LSM CFB in subjective wake after sleep onset (min) was also greater for LEM vs placebo (depression history: placebo, −47.4; LEM5, −51.8; LEM10, −52.2; all other subjects: placebo, −29.1; LEM5, −48.0; LEM10, −42.2). Overall, most treatment-emergent adverse events were mild/moderate.
Conclusion: Subjects with insomnia and a depression history may benefit from treatment of their insomnia with LEM.