We conducted a random-effects model network meta-analysis to examine differences between lemborexant and suvorexant in efficacy and safety outcomes for treating patients with insomnia. We searched Embase, MEDLINE, and CENTRAL from their inception until April/28/2020. Primary outcomes were subjective time to sleep onset (sTSO), subjective total sleep time (sTST), and subjective wake-after-sleep onset (sWASO) at week 1. Four double-blind, randomized controlled trials were identified (n = 3237; 72.4% female; mean age 58.0 years). The treatment arm consisted of lemborexant 10 mg/d (LEM10, n = 592), lemborexant 5 mg/d (LEM5, n = 589), suvorexant 20/15 mg/d (SUV20/15, n = 493), zolpidem tartrate extended release 6.25 mg/d (ZOL6.25, n = 263), and placebo (n = 1300). All active treatments outperformed placebo regarding sTSO at week 1; standardized mean differences (95% credible interval): LEM10 = -0.51 (-0.63, -0.39), LEM5 = -0.48 (-0.60, -0.36), SUV20/15 = -0.21 (-0.33, -0.10), and ZOL6.25 = -0.30 (-0.46, -0.14); sTST at week 1: LEM10 = -0.58 (-0.70, -0.45), LEM5 = -0.33 (-0.46, -0.21), SUV20/15 = -0.34 (-0.46, -0.23), and ZOL6.25 = -0.42 (-0.59, -0.25); and sWASO at week 1: LEM10 = -0.42 (-0.57, -0.28), LEM5 = -0.26 (-0.40, -0.11), SUV20/15 = -0.18 (-0.32, -0.05), and ZOL6.25 = -0.37 (-0.56, -0.18). Although no significant differences were found in discontinuation due to adverse events between each active drug and placebo, LEM10 and SUV20/15 were associated with greater somnolence compared with placebo. LEM10 had the largest effect size compared with placebo for all primary outcomes, although with a risk of somnolence.
Keywords: Insomnia; Lemborexant; Network meta-analysis; Suvorexant; Systematic review.
Copyright © 2020 Elsevier Ltd. All rights reserved.