A phase 2a randomised, double-blind, placebo-controlled, parallel-group, add-on clinical trial of ebselen (SPI-1005) as a novel treatment for mania or hypomania

Abstract

Rationale: Lithium is an effective prophylactic and anti-manic treatment in bipolar disorder; however, its use is declining through perceived poor tolerance and toxicity. Lithium inhibits inositol monophosphatase (IMPase), a probable key therapeutic mechanism. The anti-inflammatory drug, ebselen, also inhibits IMPase and appears well-tolerated and safe.

Objectives: To assess the efficacy of adjunctive ebselen in mania using the Young Mania Rating Scale (YMRS) (primary outcome) and the Altman Self-Rating Mania (ASRM) Scale and Clinical Global Impression-Severity Scale (CGI-S) among the secondary outcomes.

Methods: Randomised, double-blind, placebo-controlled, parallel-group trial conducted between October 2017 and June 2019, at Oxford Health NHS Foundation Trust. Pharmacy-controlled randomisation was computer-generated, with full allocation concealment. In/outpatients (n = 68) aged 18-70, experiencing mania or hypomania, were assigned to 3 weeks ebselen (600 mg bd) (n = 33) or placebo (n = 35). Participants received usual clinical care and psychotropic medication.

Results: Ebselen was numerically, but not statistically, superior to placebo in lowering scores on the YMRS (adjusted mean difference and 95% confidence interval, - 1.71 (- 5.34 to 1.91), p = 0.35) and ASRM (- 1.36 (- 3.75 to 1.17), p = 0.29). However, scores on the CGI-S were significantly lower at week 3 in ebselen-treated participants (adjusted mean difference, - 0.58 (- 1.14 to - 0.03), p = 0.04). A post hoc analysis excluding patients taking concomitant valproate treatment magnified the difference between ebselen and placebo on the YMRS. Adverse events were comparable between groups, and mild.

Conclusions: Ebselen merits further investigation where concomitant psychotropic medication is better controlled and participants taking valproate are excluded. If effective, ebselen's superior tolerance and safety could make it a useful alternative to lithium.

Trial registration: Trial Registry: www.clinicaltrials.gov , Identifier: NCT03013400.

Keywords: Add-on; Bipolar; Ebselen; Lithium mimetic; Mania; Randomised clinical trial; YMRS.

Fig. 1

Consort flow diagram: study sample and patient throughput. ¹In four participants, medication was ordered due to distance. However, during screening visit, n = 2, found to be ineligible; n = 1, unable to give informed consent and would not comply with trial procedures and n = 1 decided to not participate in the trial. ²As per definition of inclusion in analysis, 4 people did not start taking their medication (n = 3 withdrew consent before they started and n = 1, care team was not in agreement with patient participating in the trial)

Fig. 2

Non-adjusted mean (SEM) change from baseline scores on the Young Mania Rating Scale (YMRS) following addition of ebselen (600 mg bd) (n = 27) or placebo (n = 33) to the treatment of patients with mania/hypomania. There were no statistically significant differences at any of the time points

Fig. 3

Non-adjusted mean (SEM) change from baseline scores on the Young Mania Rating Scale (YMRS) following addition of ebselen (600 mg bd) (n = 22) or placebo (n = 27) to the treatment of patients with mania/hypomania, excluding patients taking valproate. “†” The adjusted mean difference and 95% confidence interval (CI) (change from baseline) in YMRS score between the two groups at 3 weeks was − 3.67 (− 7.39 to 0.06) (p = 0.054)

Fig. 4

Non-adjusted mean (SEM) change from baseline scores on the Altman Self Rating Scale (ASRM) following the addition of ebselen (600 mg bd) (n = 27) or placebo (n = 33) to the treatment of patients with mania/hypomania. There were no statistically significant differences at any time points

Fig. 5

Non-adjusted mean (SEM) change from baseline scores on the Clinical Global Impression Severity (CGI-S) scale following the addition of ebselen (600 mg bd) (n = 27) or placebo (n = 33) to the treatment of patients with mania/hypomania. “*” The adjusted mean difference and 95% confidence interval (CI) (change from baseline) in CGI-S score between the two groups at 3 weeks was − 0.58 (− 1.14 to − 0.03); p = 0.040