SLEEP ARCHITECTURE IN PARTICIPANTS WITH INSOMNIA DISORDER CHARACTERIZED BY PHENOTYPE: OBJECTIVE SHORT SLEEP OR OBJECTIVE NORMAL SLEEP

Background

Insomnia can be classified into 2 different phenotypes through objective sleep evaluations (insomnia with objective short-sleep duration [I-SSD]: total sleep time (TST) < 6 hours/night; insomnia with objectively longer, more normal sleep duration (INSD): TST ≥6 hours/night). I-SSD has the potential for more severe biological consequences. These analyses compared the sleep architecture responses of I-SSD and I-NSD patients to pharmacotherapy using 2 medications with different mechanisms of action, lemborexant (LEM) and zolpidem tartrate extended-release (ZOL).

Methods

Study 304 (E2006-G000-304; NCT02783729) was a 1-month, randomized, double-blind, placebo (PBO)-controlled, active-comparator study in adults (≥55 years) with insomnia disorder. Participants received PBO, LEM 5 mg (LEM5), LEM 10 mg (LEM10), or ZOL 6.25 mg. Least squares mean (LSM) duration of each sleep stage (minutes) was compared across treatment groups on Nights (NT)1/2 and NT29/30 for participants in the ISSD and I-NSD groups. Treatment-emergent adverse events were recorded.

Results

Of 1006 participants (PBO: 208; ZOL: 263; LEM5: 266; LEM10: 269), 711 (70.7%) were in the I-SSD subgroup, and 295 (29.3%) were in the I-NSD subgroup. Demographic characteristics were mostly balanced. Mean (SD) age was 64.4 (7.0) years for the I-SSD group and 62.8 (6.2) years for the I-NSD group. Males comprised 16.6% of the I-SSD group and 6.4% of the I-NSD group. In I-SSD participants, mean (SD) increases from baseline (BL) in non-rapid eye movement (non-REM) sleep time were significant for LEM5 (55.7 [37.5]; 57.6 [40.5]), LEM10 (59.9 [39.2]); 57.6 [41.6]), and ZOL (59.5 [39.8]; 47.4 [45.9]) versus PBO (19.3 [40.4]; 29.4 [41.0]) on NT1/2 and 29/30, respectively. I-SSD participants treated with LEM had significantly larger differences from BL in non-REM sleep time compared with ZOL at NT29/30. In I-NSD participants, increases from BL in non-REM sleep time were significant for LEM5 (23.2 [22.8]; 25.2 [23.4]), LEM10 (17.3 [20.7]; 16.7 [26.8]), and ZOL (26.1 [34.5]; 15.9 [30.1]) versus PBO (2.7 [25.1]; 3.1 [29.4]) on NT1/2 and 29/30, respectively. The increases from BL in non-REM sleep time were not statistically significant for LEM versus ZOL. In I-SSD participants, mean (SD) increases from BL in REM sleep time were significant at both time points, respectively, for the LEM5 (24.1 [22.0]; 18.1 [22.3]) and LEM10 (34.0 [23.1]; 24.4 [24.8]) subgroups versus the PBO (5.5 [17.4]; 5.5 [21.4]) and ZOL (9.9 [22.3]; 8.7 [22.9]) subgroups. In I-NSD participants, increases from BL in REM sleep time were significant for LEM10 (25.9 [19.9]; 15.7 [20.0]) at NT1/2 and 29/30, respectively, but only at NT29/30 for LEM5 (NT1/2: 8.6 [19.2]; NT29/30: 6.6 [17.8]) versus PBO (NT1/2: 3.8 [16.6]; NT29/30: 1.1 [19.4]). LEM5 and LEM10 led to statistically larger increases in REM sleep time versus ZOL (−4.6 [20.1]; −4.1 [18.9]) at both time points, respectively. LEM was well tolerated in both I-SSD and I-NSD subgroups.

Conclusions

Regardless of whether participants were classified as having insomnia with short sleep duration or normal sleep duration, LEM tended to increase both non-REM and REM sleep time compared with both ZOL and PBO. The magnitude of change depended on the insomnia phenotype classification, with somewhat larger increases observed in the I SSD subgroup for both drugs. Unlike LEM, ZOL only significantly increased non-REM sleep time versus placebo, likely reflecting its different mechanism of action.