Gabapentin enacarbil in Japanese patients with restless legs syndrome: a 12-week, randomized, double-blind, placebo-controlled, parallel-group study

Objective Gabapentin enacarbil (GEn) was effective and well-tolerated for the treatment of restless legs syndrome (RLS) in North American studies. However, no placebo-controlled studies of GEn have been performed in Asian patients with RLS. Therefore, we investigated the efficacy and safety of GEn in Japanese patients with RLS to determine the optimal dosage. Research design and methods Outpatients with RLS (International Restless Legs Syndrome Rating Scale (IRLS) scores >= 15) were randomized (n = 474) and treated (n = 469) in a double-blind manner with once-daily placebo (n = 116), 600 (n = 120), 900 (n = 119) or 1200 (n = 114) mg GEn for 12 weeks. Clinical trial registration NCT00530530 (ClinicalTrials.gov) Main outcome measure The primary outcome was the change in IRLS score. Secondary outcomes included Investigator (ICGI)- and Patient (PCGI)-rated Clinical Global Impression and adverse events. Results The mean change in IRLS score from baseline to the final observation was -8.96 for placebo versus -11.10, -10.28 and -11.38 for 600, 900 and 1200 mg GEn. Williams' multiple comparison test showed that only 1200 mg GEn was superior to placebo (p = 0.011). However, in post hoc mixed-effects models with repeated measures, which accounted for the time-course of changes in IRLS, the placebo-adjusted changes were -2.31, -1.92 and -2.31 for 600, 900 and 1200 mg GEn. ICGI and PCGI response rates were significantly greater for all three GEn doses versus placebo (all p <= 0.014). Adverse events, including somnolence, dizziness and nasopharyngitis, were frequent but of mild-to-moderate severity. However, there was a tendency toward a dose-dependent increase in the incidence of adverse events. Conclusions GEn is effective and well-tolerated for the treatment of RLS in Japanese patients. All three doses produced improvements in IRLS compared with placebo; 600 mg GEn is a suitable target dose. However, our analysis possibly introduced positive bias by assuming that symptoms improve after discontinuation.